The Form 10-QSB Files Dataset is a closed historical archive of every Form 10-QSB and Form 10-QSB/A submission on EDGAR from May 1994 through the discontinuation of the form in 2008, plus later amendments to pre-2008 periods. Each record is a single EDGAR accession — the unaudited small-business quarterly report required of "small business issuers" under Regulation S-B and Sections 13 or 15(d) of the Securities Exchange Act of 1934 — delivered as the primary report document, all non-image exhibits (Sarbanes-Oxley Section 302 and 906 certifications, material contracts, press releases, schedules), and a metadata.json file describing the filing at the accession level. Records are distributed as monthly ZIP containers in the original EDGAR document formats (TXT, HTML, PDF, XFD, FRM) with a JSON metadata index. The dataset is the canonical source for pre-2008 small-issuer quarterly disclosure, covering a filer population and a scaled-disclosure regime that was folded into the Smaller Reporting Company framework when Regulation S-B was repealed.
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:
One record in the Form 10QSB Files Dataset is a single EDGAR submission of Form 10-QSB or Form 10-QSB/A, identified by its 18-character SEC accession number. On disk the record is a directory whose name is the accession number with dashes stripped (accession 0001140905-08-000149 becomes folder 000114090508000149/), sitting under a year/month path (2008/2008-09/) inside a monthly ZIP archive (form-10qsb-files/2008/2008-09.zip). The directory holds every non-image document from the original EDGAR submission together with a single metadata.json that describes the filing at the accession level. A record therefore encapsulates both the regulatory artifact — an unaudited small-business quarterly report or its amendment — and the full set of certifications, exhibits, and ancillary materials that accompanied it, each preserved in its originally submitted form.
Form 10-QSB was the small-business quarterly report required under Section 13 or Section 15(d) of the Securities Exchange Act of 1934. It served as the abbreviated counterpart to Form 10-Q for issuers qualifying as "small business issuers" under Regulation S-B (generally sub-$25 million in revenue and public float) and used streamlined disclosure rules derived from Regulation S-B rather than the fuller Regulation S-K requirements. Each qualifying issuer filed one 10-QSB for each of the first three fiscal quarters, due within 45 days of quarter end. Form 10-QSB/A is an amendment to a previously filed 10-QSB; it follows the same structural template as the original and carries a cover page noting the amendment. The form was retired for periods ending on or after December 15, 2008, when the SEC repealed Regulation S-B and folded small-business quarterly reporting into Form 10-Q under the smaller reporting company framework. The dataset's temporal envelope — May 1994 through 2008 — spans the operational life of the form plus a late tail of delinquent originals and amendments filed in 2008 for much earlier reporting periods.
The dataset is distributed as monthly ZIP containers. File types observed across the corpus include TXT, JSON, HTML, XFD, PDF, and FRM — the range reflects EDGAR's modernization arc from ASCII-era paper-scan submissions to HTML-dominant modern filings.
An accession directory is organized in three strata:
metadata.json file at the root of the folder, describing the submission at the accession level. This is the canonical entry point because individual document filenames are filer-chosen and not standardized.wsrm033106q1.htm, n10qsb.htm, l10qsb123106.txt, avrg_10q.txt) and identifiable by sequence: "1" and type: "10QSB" in the metadata.EX-31.1, EX-31.2, EX-32.1, EX-32.2), and less commonly material contracts (EX-10), press releases or data schedules (EX-99), legacy financial-data schedules (EX-27), and other Regulation S-B exhibit types. Each exhibit is a separate file in the same folder.Every document file on disk retains the EDGAR SGML <DOCUMENT> envelope that would otherwise surround it inside the concatenated .txt submission wrapper on EDGAR. Each file opens with tagged header lines — <DOCUMENT>, <TYPE>, <SEQUENCE>, <FILENAME>, <DESCRIPTION>, <TEXT> — and closes with </TEXT></DOCUMENT>. Between the <TEXT> tags is whatever the filer uploaded: HTML for the HTML era, ASCII plain text for the text-era filings, or an <XFD> block for a handful of paper-scanned mid-1990s submissions.
metadata.json fieldsmetadata.json is a single JSON object per accession that mirrors the EDGAR filing-index table and adds dataset-level cross-references. Its top-level fields are:
formType — either "10QSB" or "10QSB/A".accessionNo — the dashed EDGAR accession (for example "0001140905-08-000149"); the dash-free form is the directory name on disk.filedAt — ISO-8601 timestamp with timezone offset recording EDGAR acceptance.periodOfReport — fiscal quarter end date (YYYY-MM-DD); for late or amended filings this can lag filedAt by years.description — the EDGAR form-title string, typically "Form 10QSB - Optional form for quarterly and transition reports of small business issuers".linkToFilingDetails — sec.gov URL of the primary document.linkToHtml — sec.gov URL of the EDGAR filing-index HTML page.linkToTxt — sec.gov URL of the concatenated .txt submission wrapper (referenced only; not stored locally).linkToXbrl — always empty for this dataset; 10-QSB was never designated as an XBRL-carrying filing type.documentFormatFiles[] — one entry per document in the submission, each with sequence, size (bytes), documentUrl, description, and type (EDGAR exhibit type such as 10QSB, EX-31, EX-32, EX-10, EX-99, or occasionally sparsely tagged EX-1/EX-2). A trailing entry with sequence: " " and type: " " describes the "Complete submission text file" — the concatenated .txt wrapper that lives only on EDGAR.dataFiles[] — reserved for XBRL/XML data files; always an empty array for 10-QSB.entities[] — an array of filer (and co-filer) records. Each entity carries cik, companyName (with a parenthetical role suffix such as (Filer)), irsNo, fileNo, sic (numeric code followed by a textual label, e.g. "6794 Patent Owners & Lessors"), stateOfIncorporation, fiscalYearEnd (MMDD), filmNo, act ("34" for the Exchange Act), type (the entity's role in the filing), and an optional tickers[] array.id — a 32-character hexadecimal identifier used by the sec-api service.The basename of each documentUrl in documentFormatFiles[] matches the filename on disk one-to-one, excluding the final "Complete submission" entry. This makes metadata.json the authoritative index for resolving which local file corresponds to which exhibit role, since no naming convention is enforced on the files themselves.
The primary 10-QSB document follows a standard outline that parallels Form 10-Q but with a narrower scope dictated by Regulation S-B.
The cover identifies the Commission ("SECURITIES AND EXCHANGE COMMISSION, Washington, D.C. 20549"), the form designation ("FORM 10-QSB"), the reporting period end date, two checkboxes distinguishing a quarterly report from a transition report, the exact name of the registrant, state or other jurisdiction of incorporation, IRS Employer Identification Number, principal executive office address and telephone number, securities registered under Sections 12(b) and 12(g), and the number of shares outstanding of each class of common stock as of a stated date near the filing. A final checkbox indicates whether the issuer elected the "Transitional Small Business Disclosure Format" option available under Regulation S-B.
Item numbering and presence in Part I varies across the form's lifetime and across registrants because Regulation S-B made certain items conditional. Some filings collapse Items 3 and 4 under a single "Controls and Procedures" heading; others split market-risk and controls disclosures across adjacent items.
Item numbering and titles in Part II evolved modestly over the form's life (for instance, repurchase-of-equity-securities disclosure was added as a Part II Item 2 subsection in 2004), so exact headings differ between older and newer filings.
A single signatures block at the end of the primary document containing the name of the registrant, the date, and the signatures of officers authorized to sign the report — most commonly the principal executive officer and principal financial officer, with titles spelled out. Small-business filers often have both roles held by the same person or a small handful of officers.
Exhibits are filed as separate documents in the accession folder. The exhibit set required by 10-QSB is narrower than for 10-Q, but in practice post-SOX filings almost always include:
"CERTIFICATION PURSUANT TO 18 U.S.C. SECTION 1350, AS ADOPTED PURSUANT TO SECTION 906 OF THE SARBANES-OXLEY ACT OF 2002". Each officer certifies that the report (i) fully complies with the requirements of Section 13(a) or 15(d) of the Exchange Act and (ii) fairly presents, in all material respects, the financial condition and results of operations. Signed and dated separately from the Section 302 exhibits.In the mature post-SOX regime a filing commonly carries the canonical four-certification pattern (302 CEO, 302 CFO, 906 CEO, 906 CFO). Smaller issuers frequently combine the CEO and CFO into a single document per rule, producing two certifications in total. Pre-SOX vintages omit Section 906 entirely and use an earlier Section 302 template. Some filings also carry EX-10 material contracts, EX-99 press releases or auxiliary schedules, EX-3 charter documents, and — in the oldest filings — the retired EX-27 Financial Data Schedule.
Filename conventions for exhibits are entirely filer-dependent. Observed variants for the Section 302 certification alone include ex311.htm, ex31-1.htm, ex_31-1.txt, cert_ex31z1.txt, exhibit31.htm, v126995_ex31-1.htm, and issuer-prefixed variants such as propalmsexhib311-043008.txt. Canonical identification of an exhibit's role therefore requires reading the type and sequence values in metadata.json rather than pattern-matching the filename.
metadata.json with the full accession-level metadata described above.<DOCUMENT> from the EDGAR submission — certifications, material contracts, press releases, schedules, cover letters, correspondence, and any other tagged document — each preserved under its original EDGAR filename with its SGML envelope intact.metadata.json; mid-1990s records may contain ASCII .txt documents or an .xfd paper-scan wrapper; scattered accessions include PDF cover pages, letters to shareholders, or scanned signed certifications, and FRM form-data containers appear in certain early-EDGAR workflow records.The sum of on-disk document sizes for an accession equals the size of the EDGAR "Complete submission text file" recorded in documentFormatFiles[], confirming that the per-document files on disk are precisely the pieces that the EDGAR .txt wrapper concatenates.
.jpg, .gif, .png, and any EDGAR <TYPE>GRAPHIC documents) are intentionally excluded and do not appear in any accession folder.{accession}.txt) is not stored locally. Its existence, size, and documentUrl are recorded as the final entry in documentFormatFiles[] so that a consumer can retrieve it directly from sec.gov when needed.linkToXbrl is always empty and dataFiles[] is always an empty array.Form 10-QSB was introduced by the SEC's 1992 small-business initiatives (effective 1993) and operated until its 2008 sunset. Over that span the required content shifted in several material ways:
Because 10-QSB spans EDGAR's modernization arc, the physical format of the documents inside a record varies with filing date:
<DOCUMENT> envelope, with monospaced tabular financial statements rendered using spaces and underscores. A subset of mid-1990s submissions uses the XFD paper-submission form-image wrapper for filings accepted as scanned paper documents. FRM form-data containers appear for certain EDGAR workflow records in this era.<FONT> and <TABLE> layout, frequently produced by converters such as "SEC Publisher by BCL Technologies") dominates, though a meaningful share of small-business filers continued to submit ASCII-only through the form's final years. Scattered PDF documents appear as supplemental exhibits — cover pages, letters to shareholders, scanned signed certifications.formType of 10QSB/A indicates an amendment that replaces or supplements a prior 10-QSB. The amendment carries the same structural outline as the original and restates only the affected portions, but periodOfReport refers to the original period being amended, not the amendment date. The relationship between an amendment and its original is not encoded in metadata.json beyond shared periodOfReport and cik; matching amendments to originals requires filer-level reasoning across records.filedAt of 2008-09-26 and periodOfReport of 2006-03-31 is expected rather than anomalous.type and sequence in metadata.json are the authoritative descriptors for each document's role. Filenames on disk are filer-chosen and cannot be pattern-matched reliably across the dataset.<DOCUMENT>, <TYPE>, <SEQUENCE>, <FILENAME>, <DESCRIPTION>, <TEXT>, and their closing pair). Text extraction should strip the envelope and then parse the inner body — HTML for modern filings, plain-text between <TEXT> and </TEXT> for text-era filings, or the <XFD> block for scanned paper submissions.<TABLE> elements or as space-aligned ASCII columns in text-era filings. There is no separate machine-readable financial-data layer; extraction of numeric facts requires HTML table parsing or ASCII column recognition.entities[] — the issuer — but co-filers such as subsidiary guarantors or co-registrants may appear as additional entries when the filing covers more than one reporting entity.tickers[] array in an entity record is present only when the issuer had an assignable ticker at filing time; many thinly traded small-business issuers have an empty or absent tickers[].cik reproduces the EDGAR archive path edgar/data/{cik}/{accession_no_dashes}/, enabling direct URL construction back to the authoritative source.The filer of every record is the reporting company itself (or its filing agent acting on its behalf); the report is filed, not furnished. Each accession consists of the primary quarterly report plus any exhibits, certifications, and financial schedules attached to that submission.
Form 10-QSB was restricted to small business issuers as defined in Item 10 of Regulation S-B (17 CFR 228.10). To qualify, at entry into the S-B system the issuer had to:
Once in the regime, an issuer could continue on Form 10-QSB until it exceeded either threshold for two consecutive fiscal years, after which it transitioned to Form 10-Q and Form 10-K beginning the next fiscal year.
The resulting filer population consists almost entirely of small domestic operating companies: newly public microcaps, early-stage companies that had completed SB-2 registrations, shell companies that had acquired operating subsidiaries, and long-standing small reporting companies filing under Section 13 or 15(d).
The obligation arose under Section 13(a) or Section 15(d) of the Securities Exchange Act of 1934, implemented through Rule 13a-13 and Rule 15d-13:
A qualifying small business issuer could elect Form 10-QSB in place of Form 10-Q. Use of the SB forms was elective, not mandatory, throughout the life of the regime.
Form 10-QSB was required for the first three fiscal quarters of each fiscal year. No 10-QSB was filed for the fourth quarter; those results were reported in the annual Form 10-KSB.
The deadline was a uniform 45 calendar days after quarter-end. Unlike Form 10-Q, which compresses to 40 or 35 days for accelerated filer and large-accelerated filers, Form 10-QSB had no tiered deadlines: every S-B issuer was by definition below the accelerated-filer thresholds, so the 45-day window applied across the entire filer population without variation.
A Form 10-QSB/A is an amendment to a previously filed 10-QSB, filed to correct errors, restate interim financials, respond to SEC staff comments, add missing exhibits or certifications, or reflect subsequent events or reclassifications. Amendments had no independent statutory deadline; they were filed promptly once the need to correct or supplement was identified. Each 10-QSB/A is a separate EDGAR submission with its own accession number and appears as a distinct record from the original it amends.
Content requirements came from Regulation S-B (the SEC's scaled disclosure system for small business issuers) rather than Regulations S-K and S-X. A 10-QSB contained unaudited interim financial statements under GAAP with S-B presentation rules, MD&A under Item 303 of Regulation S-B, legal proceedings and changes-in-securities disclosures, and, where applicable, quantitative and qualitative market risk disclosures.
After the Sarbanes-Oxley effective dates, 10-QSB filings also included:
These certifications appear as Exhibits 31.1, 31.2, 32.1, and 32.2 and are included in the dataset alongside the primary report.
The SEC discontinued the 10-QSB regime in its "Smaller Reporting Company Regulatory Relief and Simplification" rulemaking, Release No. 33-8876 (adopted December 19, 2007; effective February 4, 2008), which:
Under the transition rules, issuers could file on Form 10-QSB for quarterly periods ending on or before December 31, 2007, with the actual submissions permitted into early 2008 within the 45-day window. The dataset therefore has a natural cutoff: the most recent original 10-QSB filings cover fiscal quarters ending on or before December 31, 2007. Form 10-QSB/A amendments to those historical periods can still be submitted later and may appear in the dataset with post-2008 filing dates.
The quarterly reporting obligation itself traces to the 1934 Act and has been implemented via Rules 13a-13 and 15d-13 since the modern periodic reporting framework took shape. Form 10-QSB was adopted in 1992 under the SEC's "Small Business Initiatives" rulemaking (Release No. 33-6949, effective August 13, 1992), which introduced the integrated Regulation S-B system along with Forms SB-1, SB-2, 10-SB, and 10-QSB. The earliest 10-QSB filings on EDGAR in this dataset date from May 1994, reflecting the phased EDGAR mandate that rolled out electronic filing across registrant groups between 1993 and 1996. The dataset spans the full operational life of Form 10-QSB on EDGAR, from 1994 through the end-of-2007 cutoff, plus later amendments to pre-2008 periods.
Form 10-QSB is a closed historical form (filings ceased after reporting periods ending December 31, 2007), so the useful comparisons are its Regulation S-B siblings, its Regulation S-K counterparts, the Smaller Reporting Company successor filings, and structured datasets derived from the same source submissions.
Form 10-Q (standard quarterly report). Same function (unaudited interim report satisfying Section 13 or 15(d), with MD&A, market risk, legal proceedings, and SOX certifications), different regime. 10-QSB was prepared under Regulation S-B with scaled requirements (two years of comparative income statements instead of three, reduced segment and selected-data items, narrower MD&A). 10-Q follows Regulations S-K and S-X in full. A given issuer filed one or the other in a given quarter, never both. After Q4 2007, 10-QSB was eliminated and its filer base migrated to 10-Q under the Smaller Reporting Company rules, which embed scaled disclosure inside a single form.
Form 10-KSB (small-business annual report). Annual counterpart to 10-QSB, same S-B regime and same filer population. Differs by cadence and audit status: 10-KSB is annual, audited, with three-year income statements, full business description, risk factors, and executive compensation; 10-QSB is quarterly, unaudited, interim. Q4 results appear inside 10-KSB rather than as a separate 10-QSB, mirroring the 10-Q/10-K relationship. A full pre-2008 small-issuer time series requires both.
Form 10-QSB/A (amendments). Amendments are included in this dataset alongside originals. They restate prior interim periods, correct errors, add missing certifications, or respond to staff comments. EDGAR retains both originals and amendments, so users reconstructing "as-last-amended" views must reconcile on CIK and period-of-report; users needing "as-originally-filed" views should filter to 10-QSB only.
Form 8-K (current reports). Overlaps in subject matter (material events, legal proceedings, changes in financial condition) but is event-driven, not periodic. 8-Ks are triggered by enumerated items on a short deadline (typically four business days) and contain a single event, not full interim financials. Use 8-K for point-in-time event disclosure; use 10-QSB for period-end financial state, MD&A, and quarterly continuity. Small issuers in the 10-QSB era filed both.
Modern Form 10-Q Files Dataset. Covers the same quarterly reporting function for reporting periods from 2008 forward (and earlier for non-SB issuers). High content overlap; the distinction is regulatory era and filer population. Post-2008 smaller reporting company 10-Qs apply scaling through Regulation S-K accommodations, not the stand-alone S-B regime. A time series crossing the 2008 boundary for a small issuer requires both datasets and must account for shifts in line items, disclosure depth, and XBRL tagging. For post-2008-only analyses, the modern 10-Q dataset alone suffices.
Broader Regulation S-B collections. A multi-form S-B dataset bundles 10-KSB, 10-QSB, SB-1, SB-2, 10-SB, and related filings. Broader in form coverage but no deeper on quarterly content. Choose 10-QSB Files for quarterly interim reporting specifically; choose a multi-form S-B collection when the question spans the full S-B lifecycle from registration through periodic reporting.
Structured parser / XBRL-fact datasets. Parsed datasets (MD&A sections, line-item tables, XBRL facts) expose specific fields as queryable rows. 10-QSB Files delivers the raw EDGAR submission documents (primary report, exhibits, metadata) in original formats (TXT, HTML, PDF, XFD, FRM, JSON metadata), excluding images. File-level preserves everything as filed and supports custom parsing, full-text search, or training-corpus construction; structured parsers answer schema-bound queries faster but discard content outside the schema. XBRL alternatives are particularly thin here because interactive-data tagging was not broadly required during the 10-QSB era.
Regulation S-B (in force through early 2008) was a self-contained parallel regime with its own items and its own scaled financial-statement rules under Item 310. The Smaller Reporting Company framework that replaced it (effective February 4, 2008) folded scaled disclosure into Regulation S-K via item-by-item accommodations and eliminated the separate form family. Result: 10-QSB filings reflect a single coherent S-B profile across all issuers; post-2008 smaller-company 10-Qs show heterogeneous S-K scaling elections. Use 10-QSB for S-B as a historical object; use modern 10-Q for current scaled disclosure practice.
The 10-QSB Files Dataset is distinguished by four simultaneous constraints: form-type specificity (10-QSB and 10-QSB/A only), a closed historical window (May 1994 through reporting periods ending December 31, 2007), inclusion of both originals and amendments, and delivery as raw EDGAR submission files rather than parsed fields. No neighboring dataset satisfies all four at once.
Because Form 10-QSB was retired for reporting periods after 2007-12-31, this dataset is a closed historical archive of small business issuer quarterly disclosure under Regulation S-B. Its users are concentrated in long-panel research, forensic reconstruction, due diligence on defunct or reverse-merged issuers, regulatory history, and training of financial NLP systems rather than live-filing monitoring.
Faculty and doctoral researchers studying small-cap disclosure quality, earnings management, fraud detection, and the effects of the Regulation S-B regime use the corpus as the canonical pre-2008 interim record. They pull unaudited balance sheets, income statements, and cash flow tables from the primary document, extract MD&A and going-concern language, and key filings by cik, period_of_report, and filed_at from metadata.json to align with restatement, delisting, and enforcement registers.
Quant teams building survivorship-bias-free backtests on micro-cap and nano-cap universes use 10-QSB to recover fundamentals missing from commercial feeds for shell companies, reverse-merger vehicles, and pink-sheet issuers. They extract quarterly revenue, operating loss, working capital, share counts, and warrant/derivative disclosures from the primary report, then join on accession-level metadata to build panel histories of failure, uplisting, and acquisition base rates.
Forensic practitioners reconstruct quarterly histories for issuers that are now defunct, merged, or in long-running civil or criminal matters. They compare the original 10-QSB against its 10-QSB/A pair to see what was restated, re-described, or withdrawn, and they rely on the Item 3 legal proceedings section, related-party and convertible-note footnotes, and the EX-31 / EX-32 certifications to tie disclosures to named officers and specific reporting windows for damages and scienter analysis.
Enhanced due diligence analysts at banks, payment providers, and corporate investigations units trace counterparties, beneficial owners, and directors whose public-company history runs through 10-QSB filings before the issuer went dark, delisted, or went private. They mine cover-page metadata, signature blocks, auditor identity, jurisdiction of incorporation, and Item 3 litigation, and treat the presence of 10-QSB/A amendments as a flag for deeper adverse-media and enforcement review.
Teams training domain-specific financial language models and extraction pipelines treat 10-QSB as a distinct register with shorter, more variable MD&A, Regulation S-B abbreviated statement formats, and frequent going-concern and shell-company language absent from full 10-Q corpora. The HTML and TXT primary documents feed pretraining and fine-tuning, metadata.json supplies supervised labels (form type, filer, period), and the heterogeneous exhibit mix including legacy PDF, XFD, and FRM formats stress-tests parsing robustness.
Scholars of U.S. disclosure policy use the corpus to document the Regulation S-B era, the diffusion of Section 302 and 906 certifications into the small business tier, and the 2008 transition to the smaller reporting company framework. They count and sample across the closed population using form-type labels (10QSB, 10QSB/A) and period_of_report, and quote directly from certification exhibits and internal-controls disclosures to trace language evolution across the discontinuity.
Data engineering groups inside data vendors and research platforms use 10-QSB as the small-issuer counterpart to 10-Q before 2008; without it, a large share of historical small-issuer fundamentals is missing. They key on accession numbers and form-type labels from metadata.json for entity resolution, load primary documents and exhibits into document stores, and splice 10-QSB and 10-Q tables on either side of the regime change to produce unified quarterly panels.
Researchers cataloging restatements, auditor changes, and control-deficiency admissions in the small-issuer population rely on the original 10-QSB paired with its 10-QSB/A amendment to measure what changed, when, and by whom. The amendment subset, filed_at timestamps, and the explanatory notes in amended filings supply the raw material for restatement indexes and event studies around going-concern and SOX 404 disclosures.
Each user class draws on a different slice of the same record — interim financials and MD&A for fundamental and textual work, EX-31 / EX-32 certifications for governance and SOX research, 10-QSB/A pairings for restatement analysis, and metadata.json accession-level fields for panel construction and entity resolution. The dataset's value comes from its completeness over a closed regulatory era.
Because Form 10-QSB was retired for reporting periods after 2007-12-31, these use cases are framed as historical-research and long-panel workflows on a closed corpus rather than live monitoring.
Join every accession on entities[].cik and periodOfReport to assemble quarter-by-quarter fundamentals for micro-cap and nano-cap issuers that commercial feeds dropped after delisting, reverse merger, or going private. The primary 10-QSB document supplies unaudited balance sheet, income statement, and cash flow tables; metadata.json fields (sic, stateOfIncorporation, fiscalYearEnd, tickers[]) anchor entity resolution. The output is a pre-2008 small-issuer panel that neither the modern 10-Q Files Dataset nor XBRL-based feeds can reconstruct, because 10-QSB was never XBRL-tagged.
Filter formType == "10QSB/A", then match each amendment to its original by shared cik and periodOfReport to isolate the restated quarters. Diff the MD&A, financial statement notes, and going-concern language between the two records, and log the filedAt gap as the restatement lag. This supports restatement indexes, Item 4 controls-and-procedures event studies, and scienter-window reconstruction for litigation support — work that single-filing datasets or 10-Q-only corpora cannot serve for the pre-2008 small-issuer tier.
Extract EX-31.1, EX-31.2, EX-32.1, and EX-32.2 exhibits across the 2002-2008 window, identified via documentFormatFiles[].type rather than filename. Measure adoption lag from the August 2002 effective date, signature-block officer concentration (CEO/CFO held by one person is common in this population), and language drift in the five numbered Rule 13a-14 paragraphs. The closed S-B population makes this a clean natural experiment distinct from what 10-Q exhibit corpora offer.
For a target CIK that later went dark, delisted, or became a reverse-merger shell, pull the full 10-QSB and 10-QSB/A history and read Part II Item 1 legal proceedings, related-party and convertible-note footnotes, and the named officers in the EX-31/EX-32 signature blocks. Combine with cover-page jurisdiction and IRS EIN from metadata.json entities[] to tie disclosures to specific officers and windows for AML/KYC adverse-media review, civil damages work, or criminal scienter analysis.
Use the dataset as a distinct pretraining and evaluation register characterized by short, variable MD&A or "Plan of Operation" narratives, frequent going-concern and shell-company language, and heterogeneous source formats (HTML, ASCII TXT, XFD paper scans, PDF exhibits, FRM workflow containers). metadata.json provides supervised labels (formType, periodOfReport, sic) and the SGML <DOCUMENT> envelope gives per-exhibit boundaries, so models can be trained to segment, classify, and extract against a register that full 10-Q corpora do not cover.
For any small issuer active on both sides of the 2008 transition, concatenate its 10-QSB history with its subsequent smaller reporting company 10-Q filings on cik. Use periodOfReport rather than filedAt to order the series (2008 tail filings commonly report periods years earlier), and normalize for the S-B to S-K scaling shift in comparative periods, segment disclosure, and MD&A depth. The result is a continuous quarterly panel across the regime change — achievable only by combining this dataset with the modern 10-Q Files Dataset.
The Form 10QSB Files Dataset is accessible through a JSON index API for metadata discovery, a full-archive download endpoint, and per-container download URLs for individual monthly archives.
Dataset Index JSON API: https://api.sec-api.io/datasets/form-10qsb-files.json
This endpoint returns dataset-level metadata and the complete list of container files. Dataset-level fields include the dataset name and description, updatedAt, earliestSampleDate (1994-05-01), totalRecords, totalSize, formTypes covered (10QSB, 10QSB/A), containerFormat (ZIP), and fileTypes included in each ZIP (TXT, JSON, HTML, XFD, PDF, FRM). The containers array lists every monthly archive with its key, downloadUrl, size, records, and updatedAt. Polling this endpoint on a schedule is the recommended way to detect which monthly containers changed in the most recent refresh run and selectively re-download only those containers.
This endpoint does not require an API key.
1
curl https://api.sec-api.io/datasets/form-10qsb-files.json
Example response:
1
{
2
"datasetId": "1f13365b-9ae0-68ea-952e-b2fa571b24f4",
3
"datasetDownloadUrl": "https://api.sec-api.io/datasets/form-10qsb-files.zip",
4
"name": "Form 10QSB Files Dataset",
5
"updatedAt": "2026-04-14T09:48:24.348Z",
6
"earliestSampleDate": "1994-05-01",
7
"totalRecords": 393175,
8
"totalSize": 3073447782,
9
"formTypes": ["10QSB", "10QSB/A"],
10
"containerFormat": "ZIP",
11
"fileTypes": ["TXT", "JSON", "HTML", "XFD", "PDF", "FRM"],
12
"containers": [
13
{
14
"downloadUrl": "https://api.sec-api.io/datasets/form-10qsb-files/2008/2008-03.zip",
15
"key": "2008/2008-03.zip",
16
"size": 89412673,
17
"records": 1824,
18
"updatedAt": "2026-04-14T09:48:24.348Z"
19
}
20
]
21
}
Download Entire Dataset: https://api.sec-api.io/datasets/form-10qsb-files.zip?token=YOUR_API_KEY
Downloads the full dataset as a single ZIP archive containing every monthly container from the earliest sample date to the most recent refresh. This endpoint requires an API key passed as the token query parameter.
1
wget "https://api.sec-api.io/datasets/form-10qsb-files.zip?token=YOUR_API_KEY"
Download Single Container: https://api.sec-api.io/datasets/form-10qsb-files/2008/2008-03.zip?token=YOUR_API_KEY
Each container is a monthly ZIP archive addressable by its containers[].key from the index API. To build a per-month URL, append the key value (for example, 2008/2008-03.zip) to the dataset base path https://api.sec-api.io/datasets/form-10qsb-files/ and add the token query parameter. This endpoint requires an API key.
1
curl -O "https://api.sec-api.io/datasets/form-10qsb-files/2008/2008-03.zip?token=YOUR_API_KEY"
Each monthly ZIP unpacks into a single YYYY-MM/ directory. Inside that directory, one subdirectory per filing is named by its SEC accession number and contains a metadata.json describing the filing along with all associated EDGAR documents (images excluded):
1
2008-03.zip
2
└── 2008-03/
3
β”œβ”€β”€ 0001104659-08-018432/
4
β”‚ β”œβ”€β”€ metadata.json
5
β”‚ β”œβ”€β”€ primary-document.htm
6
β”‚ β”œβ”€β”€ exhibit-31-1.htm
7
β”‚ └── ...
8
β”œβ”€β”€ 0001144204-08-015673/
9
β”‚ β”œβ”€β”€ metadata.json
10
β”‚ └── ...
11
└── ...
The dataset covers Form 10-QSB and Form 10-QSB/A — the small-business quarterly report required under Section 13 or 15(d) of the Securities Exchange Act of 1934 and its amendments. Both form types are delivered as raw EDGAR submissions, with originals and amendments appearing as distinct accessions.
One record is a single EDGAR submission identified by an 18-character SEC accession number. It is stored as a directory containing a metadata.json describing the filing at the accession level, the primary 10-QSB or 10-QSB/A report, and every non-image exhibit (certifications, material contracts, press releases, schedules) preserved under its original EDGAR filename with its SGML <DOCUMENT> envelope intact.
Form 10-QSB was elected by "small business issuers" as defined in Item 10 of Regulation S-B — U.S. or Canadian issuers with both revenues and public float under $25 million, that were not investment companies, and whose parent (if majority-owned) also qualified. Use was elective, not mandatory; qualifying issuers could alternatively file Form 10-Q. Foreign private issuers (other than Canadian) and registered investment companies were ineligible.
The earliest filings are from May 1994, reflecting the phased EDGAR mandate rollout. The most recent original 10-QSB filings cover fiscal quarters ending on or before December 31, 2007, after which Regulation S-B was repealed and small-business filers migrated to Form 10-Q. Form 10-QSB/A amendments to pre-2008 periods can still appear with post-2008 filing dates.
No. Form 10-QSB was never designated as an XBRL-carrying filing type, and the XBRL mandate post-dates the form's 2008 retirement. In every record, linkToXbrl is empty and dataFiles[] is an empty array. Extraction of numeric facts requires HTML table parsing or ASCII column recognition against the primary document.
Records contain TXT, JSON, HTML, XFD, PDF, and FRM files. Modern-era records are almost entirely HTM/HTML documents plus metadata.json; mid-1990s records may contain ASCII .txt bodies or an .xfd paper-scan wrapper; scattered PDF exhibits appear as cover pages, letters to shareholders, or scanned certifications; and FRM form-data containers surface in certain early-EDGAR workflow records. Binary image exhibits are intentionally excluded.
Both cover unaudited interim quarterly reports under Section 13 or 15(d), but 10-QSB was prepared under Regulation S-B with scaled requirements (two years of comparative income statements, narrower MD&A, reduced segment and selected-data items), while 10-Q follows Regulations S-K and S-X in full. A given issuer filed one or the other in a given quarter, never both. For a small issuer active across the 2008 transition, a continuous quarterly panel requires concatenating 10-QSB Files with the modern 10-Q Files Dataset on cik, using periodOfReport to order the series.
The dataset is distributed as monthly ZIP containers. A public JSON index endpoint (https://api.sec-api.io/datasets/form-10qsb-files.json) lists every container with key, downloadUrl, size, records, and updatedAt. A full-archive ZIP and per-month ZIPs are available via token-authenticated endpoints. Each monthly ZIP unpacks to a YYYY-MM/ directory containing one subdirectory per accession.