SEC Comment Letters and Correspondence Dataset

The SEC Comment Letters and Correspondence Dataset contains the complete public record of written communications between SEC staff and registrants during the SEC's filing review process. It covers all EDGAR submissions of form types CORRESP (registrant response letters) and LETTER (SEC staff comment letters) from March 2004 through the present — approximately 600,000 files totaling ~18 GB. The dataset gives researchers, lawyers, accountants, and investors direct access to the regulator's own accounting questions, disclosure insufficiency determinations, and interpretive positions applied to specific company filings.

Update Frequency
Daily
Updated at
2026-04-05
Earliest Sample Date
2004-03-01
Total Size
19.0 GB
Total Records
615,128
Container Format
ZIP
Content Types
HTML, JPG, JSON, TXT, GIF, PDF
Form Types
CORRESP, LETTER

Dataset APIs

Programmatically retrieve the full list of dataset archive files, download URLs and dataset metadata.

Dataset Index JSON API

Download the entire dataset as a single archive file.

Download Entire Dataset:

Download a single container file (e.g. monthly archive) from the dataset.

Download Single Container:

Dataset Files

265 files · 19.0 GB
Download All
2026-03.zip352.7 KB48 records
2026-02.zip1.4 MB244 records
2026-01.zip2.3 MB315 records
2025-12.zip2.6 MB177 records
2025-11.zip989.2 KB127 records
2025-10.zip638.2 KB125 records
2025-09.zip16.5 MB2,455 records
2025-08.zip21.4 MB2,564 records
2025-07.zip31.2 MB3,043 records
2025-06.zip27.7 MB3,193 records
2025-05.zip31.0 MB2,219 records
2025-04.zip35.0 MB2,235 records
2025-03.zip69.0 MB2,401 records
2025-02.zip53.7 MB2,166 records
2025-01.zip53.9 MB2,231 records
2024-12.zip38.2 MB2,397 records
2024-11.zip46.2 MB2,403 records
2024-10.zip66.4 MB2,817 records
2024-09.zip71.4 MB3,071 records
2024-08.zip41.2 MB2,713 records
2024-07.zip129.6 MB3,110 records
2024-06.zip46.3 MB2,862 records
2024-05.zip357.1 MB4,231 records
2024-04.zip164.3 MB3,478 records
2024-03.zip71.9 MB2,427 records
2024-02.zip96.9 MB2,592 records
2024-01.zip56.1 MB2,922 records
2023-12.zip49.8 MB2,519 records
2023-11.zip34.7 MB2,546 records
2023-10.zip37.4 MB2,431 records
2023-09.zip41.4 MB2,868 records
2023-08.zip61.0 MB3,236 records
2023-07.zip56.7 MB2,630 records
2023-06.zip66.6 MB2,711 records
2023-05.zip85.8 MB2,671 records
2023-04.zip131.8 MB2,874 records
2023-03.zip32.3 MB2,519 records
2023-02.zip31.9 MB2,243 records
2023-01.zip150.1 MB2,302 records
2022-12.zip94.1 MB2,856 records
2022-11.zip59.7 MB2,738 records
2022-10.zip51.9 MB2,713 records
2022-09.zip68.9 MB3,485 records
2022-08.zip46.6 MB3,248 records
2022-07.zip81.8 MB2,672 records
2022-06.zip113.9 MB2,863 records
2022-05.zip81.7 MB2,847 records
2022-04.zip53.7 MB2,464 records
2022-03.zip83.2 MB2,114 records
2022-02.zip26.9 MB1,882 records
2022-01.zip25.4 MB2,097 records
2021-12.zip67.1 MB2,649 records
2021-11.zip33.3 MB2,743 records
2021-10.zip41.1 MB3,099 records
2021-09.zip50.6 MB3,150 records
2021-08.zip37.8 MB2,781 records
2021-07.zip38.4 MB2,988 records
2021-06.zip198.2 MB3,290 records
2021-05.zip43.0 MB2,802 records
2021-04.zip67.9 MB3,146 records
2021-03.zip62.4 MB3,449 records
2021-02.zip44.2 MB2,935 records
2021-01.zip46.7 MB2,855 records
2020-12.zip43.4 MB2,891 records
2020-11.zip34.2 MB2,380 records
2020-10.zip38.4 MB2,619 records
2020-09.zip67.6 MB2,815 records
2020-08.zip55.0 MB2,485 records
2020-07.zip43.9 MB2,626 records
2020-06.zip37.7 MB2,458 records
2020-05.zip26.1 MB1,897 records
2020-04.zip59.0 MB2,035 records
2020-03.zip37.7 MB1,762 records
2020-02.zip26.0 MB1,728 records
2020-01.zip29.2 MB1,828 records
2019-12.zip48.4 MB1,766 records
2019-11.zip32.2 MB1,887 records
2019-10.zip32.5 MB2,027 records
2019-09.zip31.2 MB2,054 records
2019-08.zip33.6 MB1,974 records
2019-07.zip39.9 MB2,177 records
2019-06.zip43.0 MB1,951 records
2019-05.zip68.8 MB2,181 records
2019-04.zip44.3 MB2,228 records
2019-03.zip43.9 MB1,527 records
2019-02.zip48.2 MB1,879 records
2019-01.zip8.0 MB402 records
2018-12.zip51.1 MB1,904 records
2018-11.zip30.0 MB1,850 records
2018-10.zip51.0 MB2,271 records
2018-09.zip53.2 MB2,151 records
2018-08.zip68.8 MB2,232 records
2018-07.zip59.6 MB2,008 records
2018-06.zip42.6 MB1,896 records
2018-05.zip66.2 MB1,990 records
2018-04.zip52.0 MB1,856 records
2018-03.zip153.0 MB1,684 records
2018-02.zip47.6 MB1,469 records
2018-01.zip62.2 MB1,756 records
2017-12.zip109.7 MB2,003 records
2017-11.zip49.4 MB1,699 records
2017-10.zip67.1 MB2,046 records
2017-09.zip132.8 MB2,112 records
2017-08.zip95.1 MB1,958 records
2017-07.zip80.8 MB1,845 records
2017-06.zip110.0 MB2,172 records
2017-05.zip88.6 MB2,056 records
2017-04.zip122.0 MB2,036 records
2017-03.zip89.8 MB1,942 records
2017-02.zip109.1 MB1,691 records
2017-01.zip80.5 MB1,925 records
2016-12.zip78.9 MB1,957 records
2016-11.zip65.8 MB1,829 records
2016-10.zip347.3 MB3,186 records
2016-09.zip111.6 MB2,334 records
2016-08.zip120.4 MB2,293 records
2016-07.zip92.9 MB2,041 records
2016-06.zip200.1 MB2,450 records
2016-05.zip118.8 MB2,402 records
2016-04.zip84.9 MB2,088 records
2016-03.zip158.4 MB2,144 records
2016-02.zip104.7 MB1,707 records
2016-01.zip72.3 MB1,650 records
2015-12.zip115.3 MB2,110 records
2015-11.zip77.5 MB1,608 records
2015-10.zip89.6 MB1,905 records
2015-09.zip101.0 MB1,963 records
2015-08.zip103.1 MB1,886 records
2015-07.zip157.2 MB2,472 records
2015-06.zip127.1 MB2,442 records
2015-05.zip105.4 MB2,060 records
2015-04.zip101.1 MB2,236 records
2015-03.zip105.1 MB1,933 records
2015-02.zip81.9 MB1,603 records
2015-01.zip86.8 MB1,864 records
2014-12.zip105.4 MB2,305 records
2014-11.zip108.9 MB1,788 records
2014-10.zip113.4 MB2,391 records
2014-09.zip113.5 MB2,471 records
2014-08.zip128.0 MB2,361 records
2014-07.zip147.1 MB2,698 records
2014-06.zip128.0 MB2,596 records
2014-05.zip98.3 MB2,250 records
2014-04.zip190.4 MB2,869 records
2014-03.zip112.3 MB2,478 records
2014-02.zip106.1 MB2,256 records
2014-01.zip109.8 MB2,647 records
2013-12.zip107.7 MB2,338 records
2013-11.zip98.1 MB2,153 records
2013-10.zip117.0 MB2,789 records
2013-09.zip139.0 MB2,683 records
2013-08.zip106.8 MB2,531 records
2013-07.zip143.4 MB3,266 records
2013-06.zip153.6 MB3,072 records
2013-05.zip126.5 MB2,791 records
2013-04.zip148.0 MB3,257 records
2013-03.zip139.6 MB2,619 records
2013-02.zip110.7 MB2,335 records
2013-01.zip197.8 MB3,157 records
2012-12.zip122.9 MB2,423 records
2012-11.zip158.1 MB2,401 records
2012-10.zip114.4 MB2,748 records
2012-09.zip117.9 MB2,507 records
2012-08.zip127.9 MB2,809 records
2012-07.zip96.9 MB2,592 records
2012-06.zip95.2 MB2,666 records
2012-05.zip81.5 MB2,498 records
2012-04.zip89.6 MB2,695 records
2012-03.zip69.4 MB2,509 records
2012-02.zip58.9 MB2,501 records
2012-01.zip132.0 MB2,699 records
2011-12.zip78.1 MB2,675 records
2011-11.zip111.8 MB2,698 records
2011-10.zip83.5 MB2,631 records
2011-09.zip92.2 MB3,084 records
2011-08.zip80.8 MB2,968 records
2011-07.zip94.4 MB3,077 records
2011-06.zip82.8 MB3,119 records
2011-05.zip62.6 MB2,735 records
2011-04.zip82.8 MB3,231 records
2011-03.zip90.8 MB3,181 records
2011-02.zip81.7 MB3,071 records
2011-01.zip95.6 MB3,126 records
2010-12.zip86.0 MB3,444 records
2010-11.zip69.9 MB2,998 records
2010-10.zip74.3 MB3,063 records
2010-09.zip121.0 MB3,389 records
2010-08.zip102.7 MB3,337 records
2010-07.zip67.7 MB3,095 records
2010-06.zip97.4 MB3,254 records
2010-05.zip76.2 MB2,975 records
2010-04.zip57.9 MB3,213 records
2010-03.zip67.5 MB3,173 records
2010-02.zip63.2 MB2,568 records
2010-01.zip79.8 MB3,072 records
2009-12.zip57.1 MB3,016 records
2009-11.zip74.0 MB2,925 records
2009-10.zip90.1 MB3,286 records
2009-09.zip103.7 MB3,588 records
2009-08.zip62.2 MB2,988 records
2009-07.zip70.2 MB3,224 records
2009-06.zip83.1 MB3,140 records
2009-05.zip67.6 MB2,895 records
2009-04.zip65.2 MB3,157 records
2009-03.zip66.5 MB3,037 records
2009-02.zip53.4 MB2,328 records
2009-01.zip46.9 MB2,386 records
2008-12.zip89.0 MB2,948 records
2008-11.zip42.2 MB2,006 records
2008-10.zip108.2 MB2,948 records
2008-09.zip60.8 MB2,597 records
2008-08.zip54.0 MB2,481 records
2008-07.zip69.7 MB2,619 records
2008-06.zip71.2 MB2,259 records
2008-05.zip41.8 MB2,048 records
2008-04.zip67.5 MB2,234 records
2008-03.zip38.8 MB2,223 records
2008-02.zip49.2 MB2,392 records
2008-01.zip58.9 MB2,731 records
2007-12.zip55.0 MB2,437 records
2007-11.zip120.8 MB2,670 records
2007-10.zip54.3 MB2,686 records
2007-09.zip52.4 MB2,481 records
2007-08.zip112.7 MB2,841 records
2007-07.zip49.6 MB2,277 records
2007-06.zip64.7 MB2,241 records
2007-05.zip45.6 MB2,253 records
2007-04.zip44.2 MB2,105 records
2007-03.zip59.8 MB2,311 records
2007-02.zip55.9 MB2,330 records
2007-01.zip53.6 MB2,471 records
2006-12.zip55.1 MB2,254 records
2006-11.zip66.6 MB2,683 records
2006-10.zip76.7 MB2,688 records
2006-09.zip71.9 MB2,586 records
2006-08.zip90.0 MB3,082 records
2006-07.zip57.5 MB2,268 records
2006-06.zip49.3 MB2,484 records
2006-05.zip90.3 MB2,324 records
2006-04.zip21.1 MB2,010 records
2006-03.zip34.6 MB2,414 records
2006-02.zip27.2 MB2,166 records
2006-01.zip19.0 MB2,184 records
2005-12.zip12.9 MB1,747 records
2005-11.zip48.5 MB1,846 records
2005-10.zip17.1 MB1,853 records
2005-09.zip27.2 MB2,136 records
2005-08.zip17.3 MB2,212 records
2005-07.zip26.3 MB2,190 records
2005-06.zip17.0 MB2,075 records
2005-05.zip9.3 MB1,559 records
2005-04.zip15.8 MB1,429 records
2005-03.zip6.4 MB1,089 records
2005-02.zip3.9 MB693 records
2005-01.zip2.5 MB501 records
2004-12.zip2.8 MB522 records
2004-11.zip4.8 MB423 records
2004-10.zip1.5 MB301 records
2004-09.zip1.4 MB172 records
2004-08.zip277.8 KB67 records
2004-07.zip46.8 KB4 records
2004-06.zip2.4 KB1 records
2004-05.zip3.1 KB2 records
2004-04.zip10.8 KB2 records
2004-03.zip8.4 KB2 records

What This Dataset Contains

The dataset captures both sides of the SEC staff review dialogue:

SEC staff comment letters (LETTER) are formal written communications from Division of Corporation Finance, Division of Investment Management, or Division of Trading and Markets staff to a registrant. Each letter identifies the filing under review, lists numbered comment items citing specific sections or disclosure elements, directs the registrant to respond within 10 business days, and identifies the reviewing staff attorney or accountant by name and title.

Registrant response letters (CORRESP) are filed by the issuer — or its outside securities counsel — in reply to an SEC staff comment letter. Each response reproduces the SEC's comment items verbatim and addresses them in order, either committing to revised disclosure, providing supplemental financial analysis, or arguing that existing disclosure is adequate.

Together, these two form types reconstruct the complete regulatory review dialogue: the SEC's questions and the registrant's answers. No other EDGAR dataset type captures this exchange. Primary filings (10-K, S-1, proxy) present the registrant's voice; comment letters and correspondence present the regulator's evaluation of those disclosures.

Content Structure of a Single Record

Each record corresponds to one EDGAR filing submission. Submissions from 2004–2010 are predominantly plain TXT. Post-2010 submissions are predominantly HTML. Recent Staff comment letters are published as PDF files and included as PDF-to-text converted TXT files.

Structure of an SEC Staff Comment Letter (LETTER)

A typical LETTER submission contains:

  1. Header block: SEC letterhead, date, addressee (CFO, General Counsel, or outside counsel), and explicit reference to the filing under review — e.g., "We have reviewed your Annual Report on Form 10-K for the fiscal year ended December 31, 2023, filed February 15, 2024, and have the following comments."
  2. Numbered comment items: Each comment cites a specific section or item in the original filing and may request expanded disclosure, question an accounting treatment, flag inconsistencies between financial statements and MD&A, ask for non-GAAP reconciliation, or request segment-level or disaggregated financial data.
  3. Response instructions: Directs the registrant to respond within 10 business days via EDGAR CORRESP and to confirm whether future filings will incorporate the requested disclosures.
  4. Staff contact: Name, direct phone, and title of the reviewing staff attorney or accountant.

Structure of a Registrant Response Letter (CORRESP)

A typical CORRESP submission contains:

  1. Header block: Company or law firm letterhead, date, addressed to the specific SEC staff member.
  2. Reference identification: Explicit reference to the SEC letter date and the accession number of the filing under review.
  3. Comment-by-comment responses: Each SEC comment is reproduced in full (typically in bold or italic) followed by the registrant's response — committing to revise disclosure, providing supplemental analysis, arguing adequacy under applicable standards, or proposing alternative disclosure language.
  4. Closing acknowledgment: Standard paragraph in which the registrant acknowledges responsibility for disclosure adequacy, that staff comments do not foreclose Commission action, and that it will not assert staff comments as a defense in proceedings.

High-Frequency Comment Topics Across the Corpus

The following topics appear with sufficient regularity across the corpus to constitute a practical taxonomy of SEC disclosure focus areas:

  • Revenue recognition (ASC 605, ASC 606 / IFRS 15) — performance obligations, variable consideration, contract costs
  • Business combinations, purchase price allocation, goodwill impairment (ASC 805, ASC 350)
  • Non-GAAP financial measures — reconciliation completeness, equal or greater GAAP prominence
  • MD&A — liquidity and capital resources, known trends, critical accounting estimates
  • Segment reporting — reportable segment identification, aggregation criteria, revenue disaggregation (ASC 280)
  • Related-party transactions — identification, materiality, arm's-length characterization (ASC 850)
  • Executive compensation in proxy statements — CD&A narrative, pay-for-performance alignment, benchmarking peer group
  • Risk factors — generic or boilerplate language, insufficient tailoring to registrant's circumstances
  • Lease accounting (ASC 842) — classification, discount rates, right-of-use asset presentation
  • Tax disclosures — uncertain tax positions (ASC 740), valuation allowances, deferred tax
  • Going concern — management assessment, mitigating conditions
  • Cybersecurity disclosures — material incident determination, board oversight (post-2023 rules)
  • Climate-related disclosures — material risk identification, Scope emissions (selectively reviewed post-2022)
  • SPAC-specific issues — warrant accounting (liability vs. equity), earnout arrangements, sponsor compensation (prominent 2020–2022)
  • Foreign private issuer disclosures — IFRS reconciliation, home-country exemptions

Who Files or Publishes It, and When

The Two Parties

SEC staff originate comment letters (LETTER). Corp Fin handles the largest volume by far, organized into industry offices (Office of Manufacturing, Office of Finance, Office of Technology, etc.) that each review a defined registrant population. Division of Investment Management reviews investment company and investment adviser filings. Division of Trading and Markets reviews broker-dealer and exchange filings. Letters are typically signed jointly by a Staff Attorney or Staff Accountant and a Branch Chief.

Registrants file response letters (CORRESP). The full range of public issuers appears: domestic operating companies of all sizes, foreign private issuers (FPIs) on Form 20-F/40-F/F-1, registered investment companies, SPACs (particularly 2020–2022), emerging growth companies in IPO review, REITs, banks, utilities, and shell companies. CORRESP submissions are sometimes drafted and submitted by outside securities counsel, whose letterhead appears on the document, though the filing appears under the registrant's CIK.

What Triggers a Comment Letter

Comment letters are not routine deadline events — they are initiated at SEC staff discretion:

  • Routine selective review: Corp Fin is mandated by Sarbanes-Oxley Act Section 408 to review each reporting company's Exchange Act filings at least once every three years. Staff apply risk-based screening to select which filings to review, factoring in market capitalization, prior comment history, restatements, auditor changes, and industry-wide accounting trends.
  • Registration statement review: Every Form S-1, F-1, S-4, and similar registration statement is subject to mandatory staff review before effectiveness. Corp Fin targets an initial comment letter within 30 calendar days of receipt.
  • Proxy statement review: DEF 14A filings are selectively reviewed, focusing on executive compensation CD&A narrative, say-on-pay analysis, and benchmarking peer group disclosure.
  • Merger and tender offer review: S-4 registration statements and Schedule TO/14D filings are systematically reviewed for merger consideration, fairness opinions, and financial projections.
  • Event-driven sweeps: Staff initiate targeted reviews following restatements, whistleblower referrals, short-seller reports, or coordinated industry sweeps — including the 2021 SPAC warrant classification sweep, the 2022 cybersecurity disclosure sweep, and ongoing China-based issuer VIE structure reviews.

Timing and Public Release Lag

For Securities Act registration statements, the SEC targets an initial comment letter within 30 calendar days of filing. Response deadlines are almost always 10 business days from the letter date, with extensions routinely available. Full review cycles run 4–16 weeks depending on form type and complexity; complex S-1 IPO reviews can extend over several months through 3–5 rounds of correspondence.

Comment letters and responses are withheld from EDGAR public release for 20 business days after the SEC formally closes the review. This creates a rolling 4–6 week lag at the leading edge of the dataset. Correspondence from recently closed reviews does not appear immediately in the dataset.

Regulatory Basis

The process rests on Securities Exchange Act of 1934 Section 13(a) (continuous reporting obligations), Securities Act Section 8 (registration review authority), Sarbanes-Oxley Act Section 408 (three-year review mandate), and SEC Release No. 33-8238 (2003), which established the public release policy for EDGAR correspondence. Comment letters are not enforcement actions and carry no formal sanctions in themselves, though unresolved issues can be a precursor to investigation referrals.

The comment letters corpus is structurally unlike every other major EDGAR dataset because it represents the regulator's evaluation of disclosures, not the disclosures themselves.

vs. Primary periodic and registration filings (Form 10-K, Form 10-Q, S-1, 20-F, DEF 14A): Primary filings are the registrant's voice — disclosures presented to investors. Comment letters are the SEC's response — evaluative judgments about whether those disclosures are adequate, accurate, and compliant. Combining both enables attribution analysis: identifying which disclosure changes between consecutive filings were SEC-directed.

vs. Registration statement amendments (S-1/A, F-1/A, S-4/A): Amendments show what changed in a disclosure; comment letters explain why the change was required. Without the comment letter, a disclosure revision in an S-1/A is visible but causally unexplained.

vs. No-action letters: No-action letters are voluntary and prospective — a company asks whether a contemplated action would trigger enforcement. Comment letters are mandatory and reactive — the SEC initiates them after a filing is made. No-action letters are published on SEC.gov in a separate database, not through EDGAR CORRESP/LETTER submissions, and address exemptive relief and rule interpretation rather than disclosure quality.

vs. SEC enforcement actions: Enforcement actions involve formal investigation, carry potential penalties, and represent rare, escalated outcomes. Comment letters represent the routine, high-volume oversight process — the vast majority of comment letter reviews are resolved without enforcement referral. In the SEC oversight pipeline, comment letters are the early-stage signal; enforcement actions are the downstream outcome.

DimensionComment Letters (CORRESP, LETTER)Primary Filings (10-K, S-1)No-Action LettersEnforcement Actions
Initiating partySEC staffRegistrantRegistrant / counselSEC Enforcement
NatureRegulatory review dialogueDisclosure to investorsAdvisory guidanceAdjudicatory proceeding
EDGAR form typesCORRESP, LETTER10-K, S-1, 8-K, etc.Not on EDGARAP, Admin Proc
Record countThousands (2004–2026)MillionsThousandsHundreds/year
Structured as dialogueYes — matched Q&ANoPartiallyNo
Reflects SEC's viewDirectly, per-filingImplicitly (by acceptance)ProspectivelyBy adjudication
Public release lag20 business days post-closureSame day as filingPublished as issuedPublished when filed

Who Uses This Dataset

The comment letters corpus attracts a different user profile than most EDGAR financial datasets. Because it captures the SEC's regulatory evaluation of disclosures rather than the disclosures themselves, the primary users are professionals whose work depends on understanding what SEC staff expects, has questioned, or has required companies to change.

Securities lawyers and outside counsel are the heaviest practical users. Before preparing a registration statement or annual report, counsel reviews prior comment letters received by the same client and by comparable issuers to identify areas likely to attract staff attention, research Corp Fin's current interpretation of specific rules, build firm-internal precedent databases, and draft preemptive disclosure language that has resolved similar comments in prior CORRESP filings.

In-house securities and disclosure counsel use comment letter data for ongoing disclosure governance — reviewing the company's own correspondence history, identifying active SEC focus areas relevant to their sector, and preparing disclosure committee meeting materials grounded in current staff expectations rather than generic checklists.

Audit partners and technical accounting teams at Big Four and national firms use the corpus to track SEC staff interpretations of US GAAP that are not yet reflected in FASB standards or formal guidance — monitoring how staff applies ASC 606, ASC 842, ASC 805, and ASC 740 in practice across a wide range of registrants, and using this to prepare audit committee briefings and update firm accounting positions.

Investment banking ECM and capital markets teams use the dataset to reduce IPO registration review risk — pulling recent LETTER filings from comparable S-1 issuers to build pre-filing checklists, estimating review timelines, and identifying business-model-specific disclosure issues the SEC has flagged for similar companies.

Academic researchers in accounting, finance, and law use the 22-year longitudinal corpus for empirical studies of SEC oversight behavior — testing whether comment receipt predicts restatements or enforcement, analyzing variation in comment intensity by auditor quality or firm size, and studying how accounting standard adoptions and regulatory events propagate into comment letter topics.

Financial journalists and investigative reporters use comment letters to surface regulatory scrutiny that companies have not publicly highlighted — finding SEC questions about transactions, accounting treatments, or disclosures that later proved problematic, and corroborating investigative reporting with contemporaneous regulatory documentary evidence.

Compliance officers and disclosure advisors use the dataset as a rolling regulatory intelligence feed — running periodic sweeps of recent LETTER filings in their industry to update disclosure checklists, benchmark their company's MD&A and risk factors, and prepare board presentations on current SEC disclosure expectations.

Hedge funds, short sellers, and activist investors use comment letter histories to identify companies with recurring SEC accounting questions, unresolved comment rounds, or staff scrutiny of the exact financial statement items the investor wants to examine. Multi-round review histories and comments touching goodwill impairment, related-party transactions, or going concern assessments are high-value signals.

M&A due diligence counsel and financial advisors review a target company's full CORRESP/LETTER history to assess disclosure quality, unresolved regulatory concerns, and accounting treatments questioned by the SEC that may affect post-acquisition financial statement presentation.

Specific Use Cases

1. IPO Registration Statement Preparation — Anticipating SEC Comments Before Filing

Securities counsel and ECM teams use the dataset before finalizing an S-1 to build a pre-filing comment risk checklist. They pull all LETTER filings from the prior 12–24 months directed at comparable-industry S-1 issuers, systematically extract comment topics and resolution approaches from matching CORRESP responses, and identify patterns — revenue recognition policy, non-GAAP metric definitions, risk factor specificity, segment disclosure, use-of-proceeds language — that consistently appear across comparable companies. The output is a disclosure review checklist specific to the issuer's sector and business model, grounded in recent SEC review experience rather than generic templates. Well-prepared filings clear in 1–2 comment rounds; unprepared filings typically run 3–5 rounds over 3–5 months.

2. Disclosure Quality Benchmarking Against Current SEC Standards

Compliance officers and disclosure advisors retrieve all LETTER filings from the prior two years directed at peer companies in the same SIC code, filtered to form type 10-K. They catalog comment items by section (MD&A, risk factors, financial statement notes, executive compensation) and identify which disclosure patterns consistently attract comment. Comparing the company's own current disclosure against those patterns identifies specific gaps — for example, if multiple peers were asked to provide more specific liquidity runway analysis, and the company's MD&A uses similarly general language. The output is an annual disclosure gap analysis grounded in current SEC comment experience, supporting the disclosure committee review with specific, citation-backed recommendations.

3. Tracking Evolving SEC Regulatory Focus Areas

Audit firm national offices and law firm regulatory practices extract LETTER filings from the most recent 12 months and apply NLP topic modeling or keyword frequency analysis to identify rising and falling comment themes. The approach has identified major shifts before formal guidance materialized: the emergence of climate-related disclosure comments in 2022 before formal rule finalization; the concentration of warrant accounting comments on SPAC filings in early 2021; the 2023–2024 surge in comments on AI-related business description language. The 2004–present time series enables distinguishing genuinely new topics from recurring cyclical themes. Output is quarterly regulatory intelligence reporting used to update client alerts, audit practice guidance, and firm-wide disclosure advisory materials.

4. M&A Due Diligence — Target Company Regulatory Disclosure History

M&A counsel and due diligence teams retrieve all CORRESP and LETTER filings for a target company's CIK from 2004 to present, reconstructing the full review history. The assessment covers: total number of review cycles; rounds per cycle; topics consistently raised across multiple reviews; instances where the registrant pushed back rather than amended; and any reviews not yet publicly released. Accounting-focused comments on revenue recognition, segment reporting, and related-party transactions receive particular scrutiny as indicators of post-acquisition financial reporting risk. Output is a structured comment letter history exhibit in the due diligence report, with flagged material recurring issues and document citations.

5. Securities Litigation Support — Contemporaneous Regulatory Knowledge

Plaintiff and defense counsel in securities class actions retrieve all LETTER and CORRESP filings for the defendant company's CIK for the relevant period, examining whether the SEC ever questioned the specific accounting treatment or disclosure at issue. A comment letter predating the class period may establish knowledge or recklessness; a review occurring without comment on the relevant item may support a reasonable-reliance defense. CORRESP responses containing confidential supplemental analysis (even if redacted in the public record) may be relevant to discovery about what information the SEC possessed. Output is a chronological record of the company's regulatory review dialogue with annotations to items intersecting the litigation claims.

6. Academic Research — Empirical Study of SEC Oversight Effectiveness

Accounting and finance researchers use the 600,000+ record, 22-year corpus for empirical studies at a scale not achievable with manually collected samples. Representative study designs: panel regressions testing whether comment receipt predicts future restatements or enforcement actions; cross-sectional analysis of comment intensity by auditor quality, firm size, or financial distress; event studies of market reaction to comment letter receipt; difference-in-differences designs around new accounting standard adoptions (ASC 606, ASC 842) or regulatory events (SPAC boom). The matched LETTER-CORRESP structure enables dialogue analysis — studying whether concession vs. pushback response strategies differ in achieving review closure.

7. Short Seller and Activist Research — Regulatory Red Flag Screening

Short-seller analysts and activist research teams build systematic screens across the comment letter corpus: identifying companies that received three or more rounds of comments in a single review cycle, received repeated comments on the same accounting topic across multiple review cycles, filed CORRESP responses defending a treatment the SEC later required to be revised, or had a comment letter directly preceding a restatement announcement. For specific investment targets, analysts read the full LETTER-CORRESP thread to extract the SEC's own framing of accounting questions — often more pointed and specific than anything in the public filings. The SEC's own words about a company's accounting provide analytically and rhetorically powerful support for short or activist theses.

Dataset Access and Download

The dataset is distributed as a single master ZIP archive (sec-comment-letters.zip) containing monthly sub-archives organized by year and month (e.g., 2026/2026-02.zip). Each monthly archive contains the EDGAR submission files for CORRESP and LETTER filings accepted in that month.

Full dataset download: https://api.sec-api.io/datasets/sec-comment-letters.zip

Content is provided as-filed from EDGAR. No NLP pre-processing, comment item extraction, or structured field parsing is applied. Users building structured analytical datasets must apply their own text processing to primary document files.

Frequently Asked Questions

How are comment letters and response letters linked in the dataset? The link is implicit, not explicit. A staff letter and its CORRESP response share the same registrant CIK. The response text cites the SEC letter's date and sometimes its accession number. Reconstructing a complete review thread requires parsing document text and applying date-sequence logic — there is no dedicated metadata field connecting a LETTER to its corresponding CORRESP.

Do the records include the full text of the letters? Yes. The primary document in each submission contains the full letter body as filed on EDGAR — not excerpts or summaries. The original HTML, TXT, or PDF format is preserved.

Are all portions of every letter publicly available? Not necessarily. Some letters contain passages subject to confidential treatment orders, which are redacted (typically shown as bracketed placeholders or omitted passages) before public EDGAR release. The dataset reflects the public EDGAR record, including any such redactions.

How current is the dataset? Comment letters are withheld from EDGAR public release for 20 business days after SEC staff formally closes the review. The most recent 4–6 weeks of completed reviews are typically not yet visible. The dataset is updated monthly.

Does the dataset distinguish SEC-authored letters from registrant responses? Yes. Form type LETTER identifies SEC staff comment letters; form type CORRESP identifies registrant responses. Both are filed under the registrant's CIK, but the form type field provides an unambiguous distinction.

Which SEC divisions are represented? Primarily Division of Corporation Finance (the dominant source), with Division of Investment Management and Division of Trading and Markets also represented. The letter body identifies the originating office and staff member.

Is this dataset useful for non-US companies? Yes. Foreign private issuers (FPIs) appear extensively in the corpus, particularly companies filing on Forms 20-F and F-1. Comment letters directed at FPIs often address IFRS reconciliation, home-country exemptions, and VIE structure disclosures relevant to China-based issuers.

Can this dataset be used to build AI/NLP models? Yes. The corpus size and consistent dialogue structure make it well-suited for NLP applications including topic classification, regulatory question extraction, comment-response matching, and training data for compliance-oriented language models.