Federal Register API: Intelligence Source Guide

The Federal Register is the daily legal newspaper of the United States government, and its API exposes every rule, proposed rule, notice and presidential document as structured JSON without a key. Sanctions designations, export controls, debarments and agency actions become law here before they appear anywhere else.

At a glance

Source	Federal Register API
Category	Corporate, Ownership & Legal Records › Legislation, Gazettes & Public Records
Homepage	https://www.federalregister.gov/
Machine interface	`https://www.federalregister.gov/api/v1/`
Format	`REST`
Access	Open — no account required
Disciplines	Legal Intelligence, Government Intelligence
Mission domains	Sanctions Evasion, Energy Security, Climate Security

Daily US federal rules, proposed rules, notices and executive orders; no-auth JSON REST API. — as catalogued in the platform’s own source registry.

The Federal Register is the official daily publication in which United States federal agencies are legally required to publish their rules, proposed rules and notices, and in which presidential documents – executive orders, proclamations, memoranda and determinations – are published. It has appeared every federal business day since 1936. Publication in it is not a press function; it is a legal act. A rule that has not been published is generally not effective, and the publication date starts the clocks that govern comment periods, effective dates and judicial review. The API, operated by the Office of the Federal Register in partnership with the Government Publishing Office, exposes the full corpus as JSON over unauthenticated HTTP. Each document record carries a document number, publication date, type, the issuing agency or agencies, title, abstract, action, the dates on which the action becomes effective or comments close, the docket identifiers and regulation identification numbers that link it to the rest of the rulemaking record, the Code of Federal Regulations parts it affects, page ranges in the printed volume, the official citation, topic terms, and links to HTML, PDF and plain-text renderings of the full text. The API supports full-text search with faceting and date filtering, retrieval of single documents, public inspection documents that are filed but not yet formally published, and endpoints covering agencies and the associated rulemaking metadata.

This is the source that tells you when something became legally true, with a citation you can put in front of a court. Every other route to United States regulatory intelligence – press releases, agency web pages, commercial trackers, news coverage – is downstream of this and is either later, less precise or unciteable. That matters most in four areas. Sanctions and export control: additions to the Commerce Department's Entity List and Unverified List, State Department debarments under the arms export regime, Treasury actions requiring rulemaking, and the executive orders that create sanctions programmes in the first place are all enacted here, and the document is the instrument rather than a report about the instrument. Corporate and regulatory exposure: enforcement actions, licence revocations, safety directives, import restrictions and product bans are published here with the affected entities named, which makes it a continuous stream of adverse-media-grade material with legal weight. Policy tracking: proposed rules with comment deadlines are the earliest formal signal of a coming regulatory change, and the docket identifiers let you follow who filed comments and what happened next. And presidential documents: the full text of executive orders and proclamations, in their authoritative form, on the day of publication. For LEGINT and GOVINT work on the United States this is the primary collection route, not a supplement.

Who publishes it, and why that matters

The Office of the Federal Register sits within the National Archives and Records Administration and shares operation of the website and API with the Government Publishing Office. The mandate is statutory – the Federal Register Act requires the publication and the government funds it as core legal infrastructure – so there is no commercial incentive shaping what appears and no realistic prospect of the corpus disappearing. The website and API were rebuilt as an open-government project and the codebase has been developed in the open, which is unusual for a federal legal publisher and has produced a notably good API by government standards: consistent field naming, sensible pagination, real full-text search, no key requirement. Two operational facts follow. First, the service is dependable but not contractually so – there is no service level agreement, and it has been affected by government shutdowns and funding lapses in the past, which is a real availability risk for anything time-critical. Second, the Office of the Federal Register is a publisher, not an author: it does not check whether an agency's rule is correct, lawful or coherent, only that it meets publication requirements. The authority of a document rests entirely with the issuing agency, and the Federal Register's own role adds format and date, not validation.

Provenance is the first question to ask of any dataset and the one most often skipped. Who collects it, what their incentive is, whether they publish a methodology, and whether they correct the record when they get something wrong all bear directly on how much weight a finding drawn from it can carry.

What a record actually contains

The fields you will be working with, what each one means, and whether it is something you can pivot on. Read the meanings carefully — more analysis is wrecked by misreading a field than by failing to find one, and a field that looks like an observation is often an inference.

Field	Type	What it means	Pivot value
`document_number`	string	The unique publication identifier for a document, in a year-and-sequence form. It is the stable key for the document across the API, the printed volume and citations by other agencies.	The full text, the PDF of record, and any later correction or withdrawal referencing the same number.
`publication_date`	timestamp	The date the document appeared in the Federal Register. This is the legally operative date for most purposes – it starts comment clocks and often triggers effectiveness – and it is not the date the agency signed the document.	The full issue for that day, and the surrounding documents from the same agency.
`type`	enum	Whether the document is a final rule, a proposed rule, a notice or a presidential document. The type determines its legal effect and is the first filter in any collection strategy, because notices and rules answer completely different questions.	The rulemaking lifecycle – a proposed rule usually has a later final rule under the same regulation identifier.
`agencies`	array	The issuing agency or agencies with their identifiers and hierarchy. Joint rulemakings list several. Agency is the most useful single facet for building a targeted feed.	All documents from that agency, and the agency's own docket and enforcement systems.
`regulation_id_number`	string	The RIN, which links every document in a single rulemaking from the regulatory agenda through proposal to final rule. It is how you follow one policy across years rather than seeing disconnected documents.	The full rulemaking history for that regulatory action, including its appearance in the unified regulatory agenda.
`docket_ids`	array	Agency docket identifiers, which are the key into the public rulemaking record – supporting studies, comments received and agency responses held in the federal docket system.	The docket itself, including who commented, which is often the more revealing intelligence.
`cfr_references`	array	Which titles and parts of the Code of Federal Regulations the document creates, amends or removes. This tells you what body of standing law is being changed, as opposed to what the document says about itself.	The current codified text of the affected regulation.
`effective_on`	timestamp	When a rule takes legal effect, which is frequently later than publication and occasionally earlier for good cause. The gap between publication and effect is the compliance window and is often the actionable fact.	none
`comments_close_on`	timestamp	Deadline for public comment on a proposed rule. It is the single most time-sensitive field in the corpus for anyone who intends to participate rather than merely observe.	The docket, where filed comments appear.
`abstract`	string	The agency's own summary of the document. Written by the agency, so it reflects how the agency wants the action characterised, which is analytically useful in itself but is not a neutral description.	none
`action`	string	A short statement of what the document does – final rule, notice of proposed rulemaking, interim final rule, correction, withdrawal, notice of availability. It distinguishes substantive action from housekeeping faster than reading the title.	none
`citation / volume / page range`	string	The formal citation in volume and page form, which is how the document is referenced in court filings, other regulations and agency correspondence.	The bound volume in the government publishing archive, which is the version of record for litigation.
`president / executive_order_number`	string	For presidential documents, the president under whom it was issued and, for executive orders, the sequential order number. Order numbers run continuously across administrations, which makes them a clean key.	Amending and revoking orders, which reference earlier order numbers explicitly.
`topics`	array	Subject index terms assigned to the document. Useful for broad thematic sweeps and unreliable for precise targeting, because assignment is inconsistent across agencies and eras.	Other documents sharing the term, with the caveat that recall is incomplete.

Coverage — and what is not in it

Coverage is the United States federal executive branch and the presidency, daily, on every federal business day. The digital corpus with full structured metadata and full text runs from 1994 onward, and the API surfaces that period natively; earlier volumes back to the 1936 founding exist in digitised form through the government publishing archive but with less structure and different access routes, so any long historical analysis needs a second collection path. What is included is anything an agency is required or chooses to publish: final rules, proposed rules, interim and direct final rules, notices of every kind – meetings, information collections, licence applications, enforcement actions, sanctions determinations, availability of environmental documents – and all presidential documents required to be published. Public inspection documents are available before formal publication, typically giving a day or more of advance sight on documents already filed, which is materially useful for time-sensitive monitoring. What is outside scope is equally important: the Federal Register is executive-branch only. Congress does not publish here, the courts do not publish here, and state and local government do not publish here. Independent agencies do publish here when their statutes require it. Update rhythm is a daily issue published each business day, with the public inspection list updated through the day, and corrections appearing as their own documents rather than as edits to the original.

Known blind spots

Absence of evidence here is not evidence of absence. These are the conditions under which Federal Register API will not show you something that is nevertheless real:

It is executive branch only. Statutes, bills, committee reports, hearings and the Congressional Record are not here, and neither are judicial opinions – so a legal picture built from this source alone is missing two of the three branches.
Publication does not mean enforcement. A rule published and effective may be unenforced, stayed by litigation, or superseded by an unpublished enforcement policy, and none of that appears in the document.
Agency press releases frequently announce an action days before the Federal Register document appears, so this source is authoritative but not always first. Treating absence today as evidence nothing happened will be wrong on exactly the fast-moving stories.
Sanctions and designation programmes are not fully represented. Some designations are made under authorities that do not require Federal Register publication of the list itself, and the operative list may live in an agency system while the Register carries only the authority.
The abstract and title are agency-written and frequently obscure the substance. A consequential change can sit inside a document titled as a technical amendment, and keyword monitoring on titles alone will miss it.
Full-text search operates over the document as published, so entity names appearing in supporting materials, appendices held in the docket rather than the Register, or incorporated-by-reference standards are not searchable here.
Coverage of the pre-1994 corpus through the API is limited, so historical regulatory research requires a different route and should not assume the same fields exist.
Withdrawn and corrected documents remain in the corpus, and nothing in a superseded document tells you it has been superseded unless you find the later document that did it.
The corpus reflects what agencies chose to characterise as publishable. Guidance documents, interpretive letters, enforcement discretion memoranda and a large body of consequential agency action are deliberately issued outside it.

Write the blind spot into the product. A statement that something “was not observed in Federal Register API” is defensible; a statement that it “did not happen” is not, and the difference is what survives cross-examination.

Access, licensing and what you may do with it

Access model: Open — no account required

There is nothing to arrange. The API accepts unauthenticated requests and returns JSON, with endpoints for document search, single-document retrieval, public inspection documents, agencies and related metadata. Search supports full-text queries with field-level filtering by agency, document type, publication date range, docket, regulation identifier and CFR part, and returns faceted counts that make it easy to size a query before running it. Results are paginated with a per-page cap, and the API can also return the same result set as CSV for direct import. Each document record includes links to the full text in several renderings, so the search response is a routing layer and the text is fetched separately. The website exposes the same corpus with saved searches and email alerting for users who want monitoring without building it. The developer documentation is the authoritative description of parameters and is worth reading rather than reverse-engineering, because the query grammar supports more precision than most users discover.

Licence

Federal Register documents are works of the United States Government and are in the public domain in the United States – there is no copyright to clear and no licence to accept. You may republish, index, redistribute and build commercial products on the corpus. Two caveats deserve stating. First, some documents incorporate material by reference – industry standards, technical specifications – that is itself copyrighted and not reproduced in the Register; the reference is public, the standard usually is not, and this catches people building compliance products. Second, the version of record for legal purposes is the published document as issued, so a derived or reformatted version you publish should point at the official citation rather than substituting for it. Attribution is good practice and, more importantly, a citation to volume and page is what makes a claim checkable. The API itself carries no separate terms beyond ordinary acceptable use.

Rate limits and fair use

No key means no quota tied to an identity, which means the etiquette matters more, not less. Keep concurrency low, add a delay between requests, send a user agent that identifies you and gives a contact, and never retry aggressively on error. The corpus updates once per business day for published documents and through the day for public inspection, so polling more than a few times daily is pure waste: a well-built collector checks the public inspection list a few times a day and the published issue once. For backfill, use date-bounded queries walking backwards rather than deep pagination into a single enormous result set, and store what you fetch so you never fetch it twice. If you need the whole corpus, work in monthly slices and expect the exercise to take hours rather than minutes. The service is small federal infrastructure serving a large audience; a badly behaved crawler is visible and unnecessary.

Licensing changes, and it changes without warning. A dataset that was free for research this year may not be free for commercial or evidential use next year. Confirm the current terms before you build a dependency on it, and record the terms you relied on alongside the data — the licence in force at the time of collection is part of the provenance.

Collecting it

How Federal Register API is actually pulled, in the order you would set it up. Prefer the bulk or export interface over per-item lookups wherever one exists: it is kinder to the publisher, faster for you, and gives a reproducible snapshot rather than a series of point-in-time answers you cannot reconstruct later.

Method	Format	Cadence	Notes
Daily published issue sweep	JSON	Once per federal business day	Pull all documents with the current publication date, filtered to the agencies and types you care about. This is the backbone of any monitoring build and it is small – a typical issue is a few hundred documents.
Public inspection watch	JSON	Several times per business day	Documents filed and available for inspection before formal publication. Gives advance sight on actions that will be legally effective shortly, which is the difference between reacting and anticipating on sanctions and enforcement.
Targeted full-text search	JSON	On demand and as saved queries	Entity names, programme names, statutory citations and technical terms across the corpus. The route for retrospective research and for building watchlists around named companies and individuals.
Rulemaking lifecycle tracking by RIN	JSON	Weekly	Follow a regulation identifier across proposal, extension, final rule and correction. Converts a stream of disconnected documents into a policy narrative with dates.
CSV export of a result set	CSV	Ad hoc	The search interface can return results as CSV, which is the fastest route to a working dataset for an analyst who does not want to write a client for a one-off question.
Bulk historical backfill	bulk	Once, then never	Date-sliced backwards walk to build a local corpus. Do this once, store it, and thereafter collect only the daily delta. Repeated backfill is the most common way people abuse this service.

Ingesting it into the platform

Every step below is idempotent and cursor-based: interrupt one and it resumes from where it stopped rather than duplicating rows or losing progress. Collection is recorded per source, so a feed that quietly stops publishing shows up as a stale timestamp instead of silently thinning your coverage.

Register the daily issue as a scheduled collection — collect.php runs the published-issue query on each federal business day and records the document count, so a day with zero documents raises an alert rather than passing silently as normal.
Separate public inspection from published records — ingest.php holds filed-but-unpublished documents in a distinct state, because their legal status differs and their document numbers and dates can change on publication. Merging the two produces records that appear to have been legally effective before they were.
Extract named entities from full text, not from abstracts — enrich.php pulls company names, individual names, addresses and identifiers from the full document text. Abstracts are agency summaries and routinely omit the names that make a document actionable – a debarment abstract may not name the debarred party.
Resolve entities against the existing graph — resolve-everything.php matches extracted names against organisations and persons already known to the platform, so a sanctions or enforcement document attaches to an existing profile rather than creating a duplicate. Matching is on normalised names plus corroborating identifiers, never on name alone.
Key the rulemaking lifecycle on RIN and docket — Documents sharing a regulation identifier or docket are linked as one action, so timeline.php can present a policy as a sequence with its comment periods and effective dates rather than as scattered publications.
Route by type into mission views — Sanctions and export control documents feed sanctions.php, enforcement notices feed financial-crime.php and org-profile.php, and presidential documents feed geopolitical.php. Routing is on agency and CFR part, both of which are structured, rather than on keyword guessing.
Set alerts on deadlines, not just on publication — alerts.php fires on comments_close_on and effective_on approaching, not only on the document appearing. The publication is the notification; the deadline is the thing that has consequences.
Preserve the citation with every derived record — Every fact the platform asserts from this source carries the document number and official citation through to export.php, so a downstream consumer can verify it against the version of record rather than trusting the platform.

Registered sources and their last-collected state are listed in sources.php, and the scheduled chain that keeps them current is in automation.php.

How it is wrong, and how to tell

Every dataset is wrong in characteristic ways. Knowing which ways is the difference between using a source and being used by one, and it is the part of source evaluation most often skipped because it is the part that takes work.

This is close to the top of the reliability scale available in open sources, and the reason is structural: the documents are legal instruments, the metadata is generated by the publication process rather than inferred, and errors are corrected through published corrections that are themselves documents. Dates, document numbers, citations, agency attributions and CFR references are dependable. Full text is the text as published. What is less dependable is anything editorial: topic terms are inconsistently assigned, abstracts vary in quality and candour by agency, and titles are frequently unhelpful. What is outside the source's control entirely is whether the agency action itself is sound, and the Register does not check. The other quality dimension worth stating is completeness against the world rather than against the corpus: the Federal Register is complete for what is published in the Federal Register, which is a subset of consequential federal action. Judge it as an authoritative record of formal publication – which it is, essentially without qualification – and never as a record of what agencies are doing.

Characteristic false positives

Signature date read as effective date: a document signed weeks earlier, published today, and effective in sixty days has three different dates, and picking the wrong one produces compliance advice that is confidently off by two months.
Correction and withdrawal blindness: the original document stays in the corpus and reads as authoritative. Unless you check for later documents referencing the same number or RIN, you will quote a rule that was withdrawn.
Entity name matching without corroboration: company names in enforcement and debarment documents are as written by the agency, with variant spellings, former names and no identifiers, so automated matching attaches actions to the wrong company at a meaningful rate.
Title-driven triage: consequential changes are routinely buried in documents titled as technical corrections or miscellaneous amendments, so a monitoring system filtering on titles misses precisely the actions that were meant to attract no attention.
Proposed treated as final: a notice of proposed rulemaking is a statement of intent that may never become law, and reporting or advising on it as though it were effective is the most common substantive error made with this corpus.
Absence read as inaction: an agency acting through guidance, enforcement discretion or a non-published mechanism leaves no trace here, so a clean Federal Register record is not evidence of an unregulated space.
Topic-term reliance: index terms have poor and inconsistent recall, so a search restricted to a topic will miss documents that plainly belong to it. Full-text search with well-chosen terms outperforms the taxonomy.
Joint rulemaking attribution: documents issued by several agencies are frequently attributed in analysis to whichever agency is listed first, which distorts any count of agency activity.

None of these make the source unusable. They make it a source that requires corroboration before an assertion built on it goes into a product, which is true of every source and admitted by few.

Ageing

Documents themselves never age – a 2011 final rule is permanently a 2011 final rule and its citation remains valid forever. What ages is the legal state the document describes. A rule can be amended, withdrawn, vacated by a court, superseded by statute or rendered moot by a subsequent executive order, and none of that changes the original document. So the ageing problem here is not stale data but stale inference: an assertion that a regulation currently requires something, based on the rule that created it, decays continuously and can be wrong within weeks. The defence is to treat the Federal Register as the record of changes and the Code of Federal Regulations as the record of current state, and to check the codified text before asserting present-day obligation. Comment deadlines and effective dates age to irrelevance on a known schedule, which makes them the easiest fields to handle: a proposed rule whose comment period closed two years ago is either a final rule now or was abandoned, and finding out which is one query on the regulation identifier.

What this source feeds

A source is only worth what it lets you conclude. These are the disciplines that collect through it, the mission domains it serves and the data points it yields — every one is a tag, so you can follow any thread from here into the rest of the library.

Collected by these intelligence disciplines

Serves these mission domains

Yields these data points

Company / Organization

How each sector uses Federal Register API

The same dataset is worked very differently depending on who you are, what authority you hold, and what you are ultimately producing. A military analyst is supporting a commander’s decision; a journalist is meeting a publication standard; an NGO caseworker is protecting a person. The records are shared — the constraints, thresholds and outputs are not.

🎖 Military and defence

The relevant streams are export control, defence trade and acquisition. Additions to and removals from export control lists, changes to the international traffic in arms regulations, debarments of defence contractors, and country-specific licensing policy changes are all published here and determine what may lawfully be transferred to whom. For force protection and partner engagement, presidential documents establishing or modifying sanctions programmes define the legal environment a mission operates in, and their publication date is the moment the environment changed. Acquisition and contracting personnel get the earliest formal notice of regulatory changes affecting procurement here, usually with a comment period during which the department's own position is being formed. It is unclassified, citable and available to coalition partners, which makes it a good shared reference in mixed environments.

🕵 National intelligence

For GOVINT and LEGINT this is a primary collection route on United States policy. The corpus gives you executive orders in full text on the day they publish, the regulatory apparatus that implements sanctions and export control, and the notices through which agencies announce determinations that are otherwise opaque. Proposed rules with comment periods are a forward indicator – they show what an agency intends before it acts, and the docket shows which industries mobilised against it. For counterintelligence and economic security work, enforcement and licensing notices name foreign entities and their US-facing structures with legal precision. The analytic caution is that the executive branch conducts a great deal of consequential business outside the Federal Register, so this source establishes what was formalised, and formalisation is a decision with its own politics.

👮 Law enforcement

Law enforcement uses this in two modes. As a legal reference: establishing what a regulation required on a given date, in an admissible form, which matters in any regulatory prosecution and in export control, environmental, financial and customs cases. And as lead generation: debarment notices, licence revocations, enforcement actions, import alerts and sanctions determinations name entities and describe conduct, giving you subjects with a documented federal adverse finding. The document number and citation are exactly the provenance a case file needs. What it will not give you is the investigative record behind the action, which lives with the agency and usually requires a request or a liaison relationship, and it will not tell you whether an action was later reversed unless you check.

🔍 Private investigation and corporate security

For corporate due diligence this is a high-signal, zero-cost adverse-findings source with legal weight that commercial screening products repackage and resell. A named party in a debarment, enforcement action, import restriction or licence revocation is a hard finding with a citation, not a media allegation. Searching a target company and its known aliases and former names across the corpus is a standard step and takes minutes. Two disciplines separate good work from bad: search variants and predecessors rather than the current registered name, and check whether an adverse action was later modified or vacated before putting it in a report. The public domain status means findings can be quoted to a client in full without licensing concerns.

📰 Journalism and OSINT media

This is where policy stories start, and the reporters who read it daily break stories the ones who wait for press releases do not. Proposed rules disclose intentions before they are announced; notices disclose determinations agencies would rather not headline; the gap between an agency's press release and its published document frequently reveals what was left out. Public inspection gives a short lead on documents already filed. For accountability reporting, the docket identifier is the door to who lobbied and what they said. The craft points are to read the document rather than the abstract, to distinguish proposed from final in the headline rather than in paragraph nine, and to cite by document number and date so the story remains checkable after the agency reorganises its website.

🌍 NGO, humanitarian and human rights

Advocacy and public interest organisations use the corpus both to monitor and to participate. Monitoring: proposed rules affecting environmental protection, immigration, labour, benefits and civil rights appear here first, with the comment deadline that determines whether an organisation can influence the outcome. Participation: the docket identifier is how you file a comment that becomes part of the record and that the agency is obliged to address, which is one of the few structured routes an outside organisation has into federal decision-making. For humanitarian work, sanctions programmes and their general licences and exemptions are defined through documents published here, and understanding exactly which humanitarian activity is authorised is a document-level question, not a policy-level one. Set alerts on deadlines rather than on publications, because the deadline is what forecloses action.

🎓 University and research

The corpus is a well-structured, complete-for-its-scope, freely licensed dataset spanning three decades of machine-readable federal regulatory activity, with agency attribution, dates, subject terms and full text. It supports quantitative work on regulatory output, agency behaviour, policy diffusion, the timing of executive action and the structure of notice-and-comment rulemaking, and the RIN and docket identifiers allow rulemakings to be followed as units rather than as documents. The methodological warnings that matter are the pre-1994 coverage boundary, the unreliability of topic terms as a classification scheme, the distinction between document counts and regulatory significance, and the systematic invisibility of guidance-based agency action, which biases any study treating the Register as a measure of total regulatory activity.

Playbook: working Federal Register API end to end

A repeatable sequence from first pull to finished product. Each phase states what you are trying to establish, not merely what to click — the objective is a defensible chain of reasoning, not a completed checklist.

Phase 1 — Fix the legal question before you search

Decide whether you need to know what changed, what the law currently is, or what an agency intends. The Federal Register answers the first well, the third partially through proposed rules, and the second badly – current law lives in the codified regulations. Analysts who come here for current state get a pile of amendments and no answer.

Phase 2 — Scope by agency and type before by keyword

Filter to the issuing agencies and document types relevant to your question first, then search within that. The corpus is large and keyword-first searching returns a mass of meeting notices and information collection requests that bury the substance. Agency and type are structured fields and cost you nothing in recall.

Phase 3 — Run the entity sweep with variants

For a named company or individual, search the full text for the current name, former names, common misspellings, transliterations and parent and subsidiary names. Agencies write names as they find them, and a search on the current registered name alone will miss a debarment issued under a predecessor. Keep every hit with its document number.

Phase 4 — Read the document, not the abstract

The abstract is the agency's characterisation and is systematically less informative than the text. Named parties, specific findings, conditions and exceptions live in the body. This step is where monitoring systems and analysts diverge, because a system that indexes abstracts will report that nothing happened.

Phase 5 — Establish the date structure of every action

For each document capture signature date if stated, publication date, effective date and any comment or compliance deadline, and treat them as four separate facts. Almost every substantive error made with this corpus is a collapse of these into one date. Where a rule has delayed compliance dates by category, capture those too.

Phase 6 — Follow the rulemaking by identifier

Take the regulation identification number and pull every document sharing it. You get the proposal, extensions, the final rule, corrections and sometimes a withdrawal, in order. This converts an isolated document into a policy history and immediately reveals whether what you are reading is still live.

Phase 7 — Cross to the docket for the politics

The docket identifier leads to the rulemaking record in the federal docket system, where comments filed by companies, trade associations and advocacy organisations are public. Who fought a rule, with what arguments, and what the agency said back is frequently more revealing than the rule. This is the step that turns regulatory monitoring into intelligence.

Phase 8 — Check whether the rule survived

Publication is not permanence. Search for later documents referencing the same identifier or CFR part, and check the codified text to see whether the amendment is actually in force. Rules are stayed, vacated by courts, delayed by incoming administrations and repealed, and the original document reads exactly the same throughout.

Phase 9 — Watch public inspection for time-critical work

If your work depends on knowing about sanctions, export control or enforcement actions as early as possible, poll the public inspection list rather than the published issue. The lead is short but real, and for compliance and market-sensitive contexts a day matters. Treat these records as provisional until publication assigns the final number and date.

Phase 10 — Build deadline alerting, not publication alerting

Configure alerts on approaching comment deadlines and effective dates for the documents you are tracking, not merely on new publications. A publication alert tells you something happened; a deadline alert tells you something is about to become impossible. Most organisations discover this after missing a comment period.

Phase 11 — Reconcile against the agency's own record

For any finding that will be relied on, confirm against the issuing agency's own publication of the same action – its enforcement database, its designation list, its licensing system. Agreement raises confidence; disagreement is itself a finding, and usually means one of the two is out of date in a way worth understanding.

Phase 12 — Cite by number and date in the product

Every assertion drawn from this source should carry the document number, publication date and official citation. It costs a line and it makes the finding permanently checkable, including after the agency redesigns its website and every URL you saved breaks – which will happen.

The platform ships this as a step-checked workflow in playbooks.php, so progress is recorded against a case rather than held in someone’s head.

What to pair it with

No single source carries a finding. These are the datasets that corroborate, extend or contradict this one — and a source that contradicts is worth more than one that agrees, because it is the only thing that will tell you when you are wrong.

Source	Relationship	What it adds
Regulations.gov	extends	The federal rulemaking docket system, holding the comments, studies and supporting materials behind the documents published in the Register. Where you find out who tried to change a rule and how.
Electronic Code of Federal Regulations	supersedes	The continuously updated codified text of federal regulations. The Register records changes; this records the resulting current state, which is what you need for any present-tense compliance question.
GovInfo / Official Gazettes	corroborates	Holds the authenticated bound volumes of the Federal Register, the statutes and the Congressional record, providing both the archival version of record and the legislative context the Register lacks.
USASpending.gov	extends	Links regulatory action to money: a debarred entity's federal award history, or the contracts implementing a programme created by an executive order.
OpenSecrets API	extends	Lobbying and campaign finance context for the interests visible in a rulemaking docket, connecting who commented on a rule with who funded the politics around it.
SAM.gov exclusions	corroborates	The government-wide list of excluded parties, which is the operational counterpart to debarment notices published in the Register and reflects current exclusion status rather than the historical act.
US Department of the Treasury sanctions programmes	corroborates	The operative designation lists for US sanctions, which the Register documents authorise but does not always enumerate. Always check the list itself for current designation status.
National Archives Federal Register information	prerequisite	The Office of the Federal Register's own explanation of what publication means legally, how documents are handled, and what the different document types signify.

Legal, ethical and operational constraints

There are almost no constraints on using this source and considerable risk in misusing it. The documents are public domain, publication is mandated, and no confidentiality attaches. The care required is in representation. Stating that an entity is subject to a restriction because a rule was published, without checking whether it remains in force, is the kind of error that produces liability for the person who published the claim rather than for the agency. Debarment, enforcement and designation notices name real people and companies, and republishing those names carries defamation and data protection considerations in jurisdictions less permissive than the United States – the fact of publication in an official gazette is a strong defence in most systems but not an automatic one, particularly where the action was later reversed and your version was not updated. For European data protection purposes, processing names from an official publication is lawful on ordinary grounds but does not exempt you from accuracy and retention obligations. In court, cite the official citation and, where the stakes justify it, the authenticated version from the government publishing archive rather than the API rendering.

Operational security

Requests are unauthenticated and go to federal infrastructure, so the exposure is the query itself. A daily sweep of an agency's output is unremarkable and indistinguishable from the very large volume of legitimate compliance and research traffic. A full-text search for a specific company or individual is a different matter: it states who you are interested in, and repeated targeted searches over time sketch an investigation. The website's saved search and alert features require an account and therefore attach your interests to an identity and an email address, which is convenient and is a disclosure. The safer pattern for sensitive work is to collect broadly through the API into a local store and search locally, so that the only thing visible externally is that you take a daily copy of a public gazette – which reveals nothing. Where the target is a foreign entity and the work is sensitive, the same reasoning applies with more force.

Two rules that hold regardless of jurisdiction. Collection that is lawful is not automatically proportionate, and a dataset assembled for one purpose does not carry consent for another. Where the records concern identifiable people, the question is not only whether you may hold the data but whether holding it serves the purpose you are accountable for.

Is it earning its place?

Sources accumulate. Feeds get added during an incident and are never reviewed again, and a decade later the pipeline is carrying dead weight that nobody dares remove. These are the measures that show whether Federal Register API is contributing anything, and they are worth baselining now so the answer is available later.

Latency between a document's public inspection filing and its arrival in your platform, which is the honest measure of whether your monitoring is early or merely present.
Proportion of documents in your watched agency and type set that are triaged rather than accumulated, because an unread feed is a cost with no benefit.
Number of entities in your graph enriched with a Federal Register adverse finding, and how many of those were not already known from commercial screening – the second number is the source's actual contribution.
False attachment rate in entity matching, sampled by manual review, since name-only matching against agency-written names is the main quality risk in an automated pipeline.
Count of comment deadlines and effective dates surfaced in advance versus discovered after the fact, which measures whether the alerting is built on the right fields.
Share of published findings carrying a document number and official citation, tracked as a discipline metric.
Number of times a rulemaking-lifecycle check revealed that a document being relied on had been withdrawn, corrected or superseded, which is a direct measure of the value of that step.

Beware of volume. Indicator counts rise easily and say almost nothing. Unique contribution — findings this source produced that no other source in your stack would have — is the measure that matters, and it is usually far lower than anyone expects.

Tradecraft notes

The distinctions that separate a competent analyst from a fast one:

Four dates, not one. Signature, publication, effective and compliance dates are distinct and routinely differ by months. Record all of them and say which one you mean in every sentence that has a date in it.
The abstract is advocacy. Agencies characterise their own actions, and the characterisation is data about the agency rather than a summary of the document. Read the body for named parties, exceptions and conditions.
Proposed is not final and interim final is neither. Learn the document types and what each actually does legally, because the difference between them is the difference between a plan and an obligation.
Names in agency documents are unnormalised. Search predecessors, transliterations and misspellings, and never let an automated matcher attach an enforcement action to a company on name similarity alone.
A rule you find is not a rule in force. Check the codified text and search for later documents on the same identifier before asserting current obligation, because vacaturs and delays are common and invisible in the original.
Use the docket, not just the document. The comment record is where the affected interests reveal themselves, and it is frequently the most intelligence-dense part of a rulemaking.
Absence of publication is weak evidence. A great deal of consequential federal action happens through guidance, enforcement discretion and non-published mechanisms, and a clean Register record proves only that nothing was formally published.
Do not build monitoring on topic terms. Their assignment is inconsistent enough that recall is unreliable, and full-text search with carefully chosen phrases performs substantially better.
Cite by document number and citation rather than by URL. Government websites are reorganised, links rot, and the citation is permanent and resolvable in any law library.

Questions analysts actually ask

Is this the fastest way to learn about a new sanctions designation?

Not always. Treasury and Commerce frequently announce designations by press release and update their operative lists the same day, with the Federal Register document following. The Register is the authoritative legal instrument and the citable record, and public inspection narrows the gap, but for pure speed on designations you should watch the agency's own list. Use both: the list for currency, the Register for authority.

How far back does the API go?

The structured, full-text digital corpus runs from 1994 onward and that is what the API serves well. Earlier volumes back to 1936 exist in digitised form through the government publishing archive but with different structure and access. Any historical study crossing that boundary needs two collection routes and should not assume field parity between them.

What is the difference between the Federal Register and the Code of Federal Regulations?

The Register is the record of changes, published daily and permanent; the Code is the current codified body of regulation, organised by subject and continuously updated. A rule in the Register amends the Code. If you want to know what the law requires today, read the Code; if you want to know when and how it came to require that, and who did it, read the Register.

Do I need an API key?

No. The Federal Register API is unauthenticated and free. That makes courteous rate behaviour your responsibility rather than the server's – keep concurrency low, identify yourself in the user agent, and do not poll a daily publication every minute. Note that other federal APIs you may use alongside it, including the government publishing archive's, do require keys.

Can I rely on the abstract for automated triage?

No, and this is the most common design flaw in monitoring systems built on this corpus. Abstracts are agency-authored summaries that routinely omit named parties, specific findings and the substance of what changed. Index the full text. The additional storage and processing cost is trivial compared to the cost of a system that quietly does not see debarments.

How do I know whether a rule I found is still in effect?

Search for later documents sharing its regulation identification number or amending the same CFR parts, then check the codified text in the current Code of Federal Regulations. Rules are delayed, stayed by courts, withdrawn by incoming administrations and repealed, and the original document never changes to reflect any of that. This check takes two minutes and prevents the most embarrassing category of error.

What is public inspection and should I be collecting it?

It is the set of documents filed with the Office of the Federal Register and available for viewing before formal publication, typically giving a day or so of advance notice. Collect it if your work is time-sensitive – sanctions, export control, market-affecting enforcement. Handle it as provisional: document numbers and publication dates are assigned at publication, so records must be reconciled rather than duplicated when the document formally appears.

Does this cover independent agencies and the intelligence community?

Independent regulatory agencies publish here where their statutes require it, and many do so extensively. The intelligence community publishes very little, and where it does the documents are administrative. This is not a source of insight into classified activity, and its silence on a subject carries no information about whether activity exists.

Can I republish Federal Register content commercially?

Yes. The documents are public domain in the United States and there is no licence to accept. The practical cautions are to keep republished adverse findings current, since publishing a debarment that has since been lifted creates exposure for you rather than the agency, and to be aware that standards incorporated by reference into a rule are usually copyrighted by the standards body and are not part of the public domain text.

Standards, formats and interoperability

What this source speaks natively, and what it has to be translated into before a partner can consume it. Work that arrives in a recognised format is easier to defend, easier to hand over and easier to automate against:

The Federal Register Act and the Administrative Procedure Act, which define what must be published, what notice-and-comment requires, and what legal effect publication carries.
Regulation Identification Numbers, the identifier tying every document in a rulemaking together from the regulatory agenda through to the final rule.
Code of Federal Regulations title and part citation, the structured reference to the codified law a document amends.
Federal Register volume and page citation, the permanent legal reference used in court filings and by other agencies.
Federal agency identifiers and hierarchy, which the API exposes as structured entities and which make agency-level analysis reliable.
JSON and CSV as delivery formats, with full text available in HTML, PDF and plain text renderings of the same document.
The federal docket identifier scheme, which links Register documents to the supporting rulemaking record held in the docket system.

References

Primary documentation and authoritative references for this source. Publishers revise and retire material, so treat the retrieval date as part of the citation and re-check before relying on any of it in a formal product.

Federal Register — Office of the Federal Register, National Archives and Records Administration. The publication itself, with browse, search, saved searches and email alerting. The place to start for anyone who wants monitoring without building a collector.
Federal Register API documentation — Office of the Federal Register. The authoritative reference for endpoints, parameters, filters and response fields. Read it properly – the query grammar supports far more precision than most integrations use.
Federal Register developer resources — Office of the Federal Register. Practical guidance on building against the API, including bulk access and the relationship between the Register, the codified regulations and the docket system.
About the Federal Register — National Archives and Records Administration. The Office of the Federal Register's institutional explanation of publication requirements, document types and legal effect. Necessary reading before you assert what a document does.
Electronic Code of Federal Regulations — Office of the Federal Register / GPO. The continuously updated codified regulations. The correct source for any present-tense question about what the law requires, and the check that tells you whether a rule survived.
Regulations.gov — US General Services Administration. The federal rulemaking docket system holding comments, studies and agency responses. The route from a published rule to the interests that tried to shape it.
Regulations.gov API — US General Services Administration. Programmatic access to dockets and comments, which is how you automate the step from Register document to rulemaking record.
GovInfo — US Government Publishing Office. The authenticated archive of Federal Register volumes and the wider corpus of federal publications, including the pre-1994 material the API does not serve well.
SAM.gov — US General Services Administration. Entity registration and the government-wide exclusions list, which shows current exclusion status where the Register shows the act that created it.
Office of Foreign Assets Control — US Department of the Treasury. The operative US sanctions lists and programme documentation. Check designation status here; use the Register for the legal instrument and the citation.
Bureau of Industry and Security — US Department of Commerce. Export control administration, whose Entity List and licensing policy changes are enacted through Federal Register rules and whose own guidance explains their operational effect.
Unified Agenda of Regulatory and Deregulatory Actions — US Office of Management and Budget. The forward-looking list of planned agency rulemakings keyed on the same regulation identifiers, which is how you see an action coming before any document is published.

Link integrity: every reference above was verified with a live request when this page was generated. Where a publisher had moved or withdrawn a document, the link was repointed at a preserved copy in the Internet Archive and marked as archived. Anything with no reachable copy anywhere had its link removed rather than left to rot — the source is still credited, it simply cannot be linked.

Put it into practice

The Quantus Intel threat intelligence platform operationalises this source: it sweeps the daily issue and the public inspection list, extracts named entities from full text rather than abstracts, links documents into rulemaking lifecycles by identifier, and alerts on effective dates and comment deadlines rather than merely on publication.. Browse the full source catalogue, or follow any tag above into the rest of the library.