GeminIQ
Subscribe

EDGAR Accession Numbers and Filing URL Structure

Chad Hartman

By

Published · Last updated

An accession number looks like a string of digits with two hyphens in it — 0000320193-23-000106 — the kind of string a reader's eye slides past on the way to the actual filing. Underneath that unremarkable format, it's the single identifier that resolves to the exact URL of a specific filing, every document inside it, and every rendered XBRL table the SEC generated from it. Knowing how to turn one into a working URL is the difference between citing "Apple's 2023 10-K" and citing the one document that number actually points to.

This guide covers how EDGAR's Archives URLs are built from a CIK — the Central Index Key, the permanent numeric identifier the SEC assigns to every filer — and an accession number, where the filing index page and the primary document live inside that structure, and how the R-files behind the XBRL viewer are organized. It also covers the handful of formatting differences — hyphens present here, stripped there, CIK padding required in one place and optional in another — that break a manually constructed URL more often than anything else.

Start your 7-day free trial →

Table of Contents

What an Accession Number Is

An accession number is the unique identifier EDGAR assigns to a single submission — one specific filing of any form type, submitted on a specific date. It has three parts, written with hyphens: the 10-digit CIK of the filer or filing agent that transmitted the submission, a two-digit year, and a sequence number for that submitter within the year.

Part Position in 0000320193-23-000106 What it is Worked example
CIK of the transmitter 0000320193 The 10-digit Central Index Key of the filer or filing agent that submitted it — zero-padded 0000320193 = Apple
Two-digit year 23 The calendar year EDGAR accepted the submission 23 = 2023
Sequence number 000106 That submitter's submission count for the year 000106 = the 106th submission that filer transmitted in 2023

A single accession number maps to one complete submission, which typically bundles the primary document together with every exhibit and, for a periodic report, the full set of XBRL data behind it. That specificity is what makes an accession number more precise than a company name and a filing date. Two 10-Ks filed by the same company in the same year, or a 10-K and a later 10-K/A covering the same fiscal year, each get their own accession number, and only the exact number resolves to the exact filing.

The Archives URL Pattern: From Accession Number to Filing

Every filing lives at a URL built from three pieces: the base path, the filer's CIK, and the accession number with its hyphens removed:

https://www.sec.gov/Archives/edgar/data/{CIK}/{accession-no-dashes}/

That directory is where every document belonging to the submission lives — the primary filing, every exhibit, and the XBRL viewer files, all as separate files inside the same folder. The CIK in this path can be entered with or without its leading zeros; the Archives system doesn't require the strict 10-digit padding the XBRL API does, which is one of the few places EDGAR is more forgiving than it looks.

The Filing Index Page

Appending the accession number back in — with its hyphens restored — plus -index.htm returns a human-readable page listing every document in the submission, each one linked:

https://www.sec.gov/Archives/edgar/data/{CIK}/{accession-with-dashes}/{accession-with-dashes}-index.htm

This is the single most useful URL in the entire structure, and it's the pattern GeminIQ uses to cite every filing referenced across the platform. It requires no guessing about the primary document's filename and no knowledge of how many exhibits the filing carries. The index page lists all of it, with a direct link to each document, in one page load.

Finding the Primary Document

The primary document itself doesn't follow a predictable filename pattern the way the folder structure does. A company's 10-K might be named aapl-20230930.htm, following a ticker-plus-period-end convention, or something entirely different depending on the filer's own document-naming choices — there's no rule that guarantees the filename, only that the index page always lists it. Tesla's FY2021 10-K, for example, lives at:

https://www.sec.gov/Archives/edgar/data/1318605/000095017022000796/tsla-20211231.htm

This follows the ticker-plus-date pattern, but it's convention, not requirement. Treating the index page as the starting point, rather than guessing the primary document's filename directly, is the more reliable approach for any filing you haven't seen before.

The R-Files: Where the XBRL Viewer Tables Live

Every periodic filing's directory also contains a set of files named R1.htm, R2.htm, R3.htm, and so on — one per rendered table the SEC's XBRL viewer generated from the filing's tagged data. These are the same tables a person sees when they open a filing's XBRL viewer directly on EDGAR: the balance sheet, the income statement, the cash flow statement, and each individual footnote table. Each one is broken out as its own numbered file in the same accession directory as the primary document and exhibits. They sit alongside a MetaLinks.json file and a FilingSummary.xml file, which together map which R file corresponds to which financial statement or footnote — useful context if you're trying to jump directly to, say, the debt footnote table rather than paging through the primary document manually.

CIK Padding: Where It Matters and Where It Doesn't

CIK padding is the most common way a manually built EDGAR URL breaks, and the rule is inconsistent across the two URL systems investors are most likely to use. In Archives browse URLs — /Archives/edgar/data/{CIK}/... — the CIK can be entered with or without leading zeros; both 1318605 and 0001318605 resolve to the same company. In the XBRL data API — data.sec.gov/api/xbrl/companyconcept/CIK{CIK}/... — the CIK must be exactly 10 digits, zero-padded, with the literal string CIK prefixed directly onto the number. Passing an unpadded CIK to the API returns a 404 rather than falling back gracefully, which is the opposite of how the Archives URLs behave. Handling that padding in code is covered in How to Pull SEC Financial Data With Python.

Full-Text Search Indexes Files, Not Filings

EDGAR's full-text search doesn't index at the accession level. It indexes individual files within an accession. A single 10-K submission can produce several separate search hits: the primary annual report document, an EX-23 auditor consent, an EX-10 material contract, and even individual XBRL viewer fragments, each indexed as its own searchable unit tied to the same accession number but a different filename. A full-text search result pointing to a specific exhibit rather than the primary 10-K document isn't unusual — it means the search term appeared in that exhibit specifically, not necessarily anywhere in the primary filing.

Bulk Access: The Quarterly Full-Index Files

For anyone pulling filing lists across many companies rather than looking up filings one at a time, EDGAR publishes quarterly index files rather than requiring individual lookups:

https://www.sec.gov/Archives/edgar/full-index/{YYYY}/QTR{1-4}/{form|company|master|xbrl}.idx

Each index lists every filing submitted that quarter — company name, form type, CIK, filing date, and the file path — sorted differently depending on which of the four index types is requested. form.idx sorts by form type, company.idx sorts by company name, and master.idx combines both in filing order. For a job that needs "every 8-K filed in Q2 2026," pulling one quarterly index file is a single request in place of thousands of individual filing lookups.

Citing a Filing Precisely on GeminIQ

Every figure on GeminIQ traces back to the exact filing it came from, and the accession number is the mechanism that makes that traceable. When GeminIQ's Financial Statements view shows a number sourced from a specific 10-K or 10-Q, the underlying citation is built from precisely the pattern this guide covers: CIK plus accession number, resolved to the filing index page where the original document sits. That's a meaningfully different guarantee than a platform that cites "the company's most recent annual report" without a specific accession number attached — the difference between a citation you can click through to the exact document, and one you have to take on faith.

Everything above is the addressing scheme, not the research step built on top of it. Actually walking a figure back to the filing it came from is its own workflow, covered in Verify Financial Data Against the Original SEC Filing. Filing-navigation tools sit in between: a document-reader layer rebuilds EDGAR's document tree into a cleaner reading interface but still hands you the document, while GeminIQ starts from the tagged data and keeps the accession number attached to every value it derives from it.

Frequently Asked Questions

What are the three parts of an accession number?

The filer or filing agent's 10-digit CIK, a two-digit year, and a sequence number for that submitter within the year — written as CIK-YY-SEQUENCE, for example 0000320193-23-000106.

Do EDGAR URLs use hyphens in the accession number or not?

Both, in different places. The folder in the Archives path strips the hyphens (000032019323000106), but the index page filename restores them (0000320193-23-000106-index.htm).

Why does a CIK work without leading zeros in one EDGAR URL but not another?

Archives browse URLs accept a CIK with or without leading zeros. The XBRL data API requires the CIK zero-padded to exactly 10 digits with no exceptions, and returns a 404 rather than resolving an unpadded number.

What are the R1.htm, R2.htm files in a filing's directory?

They're the individual rendered tables the SEC's XBRL viewer generates from a filing's tagged data — one file per financial statement or footnote table, sitting in the same accession directory as the primary document and exhibits.

The accession number itself is just three numbers separated by two hyphens. What it unlocks — the exact filing, the exact document, the exact table — is the entire reason EDGAR citations can be more precise than any other kind of financial sourcing.

Start your 7-day free trial →

Wall Street's data. Main Street's price.

Institutional terminals charge thousands a year for as-filed accuracy. GeminIQ gives you the same thing for a fraction of the cost: financials built directly from raw SEC EDGAR filings, not third-party APIs, with full XBRL traceability back to the original 10-K or 10-Q. No normalized guesswork, just calculated metrics, charts, screeners, and watchlists built on numbers exactly as the company reported them. Start researching now at GeminIQ.com.

Disclaimer: The content in this blog is for educational and entertainment purposes only and does not constitute financial, legal, or tax advice. Investing involves risk, including the loss of principal. The views expressed are my own and not intended as financial advice or a guarantee of future performance.