Healthcare / DOCUMENTED CSV

Drugs@FDA Approval & Submission Reference

Research FDA applications through their recorded submissions, status dates and supporting document links. Prepared relationships, readable class definitions and explicit source-quality flags.

Fixed ACD source capture: October 16, 2026 · not a complete regulatory event log

194,160submission records

19main-table fields

29,391application records

81,281document records

01 / KNOW THE FIT

A useful join starts
with a clear meaning.

What you can do with it

Add recorded submission history to a pharmaceutical research portfolio, inspect original and supplemental records, and locate FDA document links. Application and drug-product context remain separate from submission details.

What one row means

One main-table row is a source application × submission type × submission number, with its recorded status, status date, class and related-record counts. It is not a complete log of every status transition.

Unique key
submission_key

How to join it

Use six-digit application numbers as text. submission_key connects to documents and action categories; aggregate these separate many-row tables before combining counts. Drug products use application plus three-digit product number. These identifiers are not NDCs.

Research FDA application history in Python →

Coverage, with the edges visible

All 194,160 source submission rows are retained, alongside 29,391 applications, 51,889 drug products and 81,281 document records. All 966,972 rows across the 12 original source tables have an explicit outcome; four source tables are outside this product’s scope.

Keep in mind

5,377 submission rows lack a matching application record; type and sponsor stay blank. Related records are retained and flagged, rather than dropped.
FDA covers approved human drugs within its stated scope, with incomplete older document coverage. This is not every regulated product, a clinical decision tool or live monitoring.
Status date is not receipt date or elapsed review time. Current snapshot sponsors/types are not historical values. Application-level documents are distinguished from submission attachments.
Document URLs are supplied as source links, not downloaded PDFs. Malformed or whitespace-containing links require review; reachability and PDF contents were not audited.

WHAT MAKES IT ACTUALLY CLEAN

Decodes source text, preserves identifiers, resolves FDA lookup labels, checks relationship counts and retains missing-parent records. Separates suspicious 1900 dates from analytical dates, documents one specific extra-tab repair, and quarantines three all-zero identifiers. Raw-source acceptance and offline rebuilds verify the result. How we prepare and version our data →

02 / TRY THE SHAPE

Real rows. Before you buy.

These are the first three rows from the delivered CSV, showing 6 of 19 fields. The free download contains the first five rows and all 19 fields; it is a format sample, not a coverage test.

Drugs@FDA Approval & Submission Reference · first three rows
application_number	submission_type	submission_number	submission_status	status_date	sponsor_name
000004	ORIG	1	AP	1969-07-16	PHARMICS
000004	SUPPL	10	AP	1980-05-08	PHARMICS
000004	SUPPL	11	AP	1987-05-26	PHARMICS

03 / NO GUESSING

Main table fields, explained.

The main table’s field dictionary from the current download. CSV has no built-in types; these are the intended import types.

All 19 fields in CSV order
Field	Type	Meaning
`application_number`	string	Six-digit FDA application number, imported as text. Not an NDC. Links to applications.csv when present.
`application_type`	string	NDA, ANDA or BLA from this snapshot of Applications.txt; blank if the application record is absent. Not necessarily its historical type at the action date.
`sponsor_name`	string	Sponsor in this application snapshot, outer whitespace trimmed. Not a historical sponsor at each submission date.
`application_match_status`	string	matched or missing_application_record. Missing parents are retained without guessed type or sponsor.
`submission_key`	string	ACD key: six-digit application number + colon + trimmed submission type + colon + original submission number. Example 202155:ORIG:1.
`submission_type`	string	FDA ORIG (original), SUPPL (supplement), or APPL for application-level documents; outer whitespace trimmed.
`submission_number`	string	FDA submission number within application and type; retained as text without added padding.
`submission_status`	string	Original FDA status code, including AP, TA, CR or blank. Preserved without treating every record as an approval.
`status_date`	string	YYYY-MM-DD from SubmissionStatusDate, the date the supplied status was applied. Not the submission-received date, elapsed review time, or complete event history.
`status_date_status`	string	exact_date, missing, or source_1900_date_requires_review. Blank analytical date for missing values and the suspicious source 1900-01-01 date; original retained.
`source_status_date`	string	Original source timestamp, outer whitespace trimmed; source timezone unspecified.
`submission_class_id`	string	FDA SubmissionClassCodeID; blank if not supplied.
`submission_class_code`	string	Code from the exact FDA submission-class lookup; blank if no class ID.
`submission_class_description`	string	FDA lookup description for that class; not a conclusion about therapeutic value.
`review_priority`	string	Original ReviewPriority text, including blank/UNKNOWN/N/A and other source values; not imputed.
`public_notes`	string	Source SubmissionsPublicNotes, outer whitespace trimmed; blank means no text supplied.
`document_count`	string	Number of document ID records matching application, submission type and submission number. Multiple documents may have the same URL.
`action_category_record_count`	string	Number of linked source action-category rows, including repeated categories with distinct source IDs; not distinct regulatory events.
`source_row`	string	One-based source data row, excluding header; identifies the original row within the table named in source_row_status.csv.

04 / INSIDE THE DOWNLOAD

Data, with its context.

One ZIP · 13.6 MB · 21 files. No account or subscription required.

README.mdImport guidance, method, attribution, and limitations
acceptance.jsonIndependent comparisons against original source values and relationships
all_tables_dictionary.jsonField types, meanings and join guidance for all eight CSV tables
applications.csvOne record per normalized application number, with original number, type, sponsor and counts
checksums.jsonSHA-256 checksums for the release files
code_definitions.csvComplete source submission-class, document-type and action-category lookups
data_dictionary.jsonTypes and definitions for every field
documents.csvDocument IDs, submission/application relationships, source URLs and explicit quality flags
drug_products.csvApplication/product context, with source names, ingredients, forms and strengths
example.pyCopyable Python application-portfolio example, also runnable in a notebook
example_input.csvFictional project labels paired with genuine FDA application numbers
example_output.txtExpected printed output
example_result.csvExpected application-portfolio results
metadata.jsonSource provenance, production time, counts, and CSV checksum
provenance.jsonRetained source identifiers, URLs and file checksums
quality_report.jsonCoverage, missingness and product-specific quality checks
revision.jsonChanges and free-upgrade eligibility for this ACD revision
source_exceptions.csvThree all-zero application identifiers, quarantined with original source cells
source_row_status.csvDisposition of every original row, including out-of-scope tables and quarantined records
submission_actions.csvSource action-category rows with lookup descriptions; not independently dated events
submissions.csvMain prepared table · UTF-8 CSV

05 / DATES THAT MEAN SOMETHING

Sources, versions & changes.

Source versions identify the inputs. Our processing revision describes how we prepare them. Either kind of change creates a new ACD release; the package date records when it was built.

Current source versions

ACD source capture: October 16, 2026
FDA download-page last-updated label: October 16, 2026
ZIP member timestamps: October 15, 2026; file metadata, not coverage cutoff
Exhaustive source coverage cutoff: Not stated
ACD preparation revision: 1 · related tables, source accounting and explicit quality flags
Package prepared (UTC): Oct 16, 2026

Source: U.S. Food and Drug Administration. Independent ACD preparation; no FDA endorsement. The original government tables remain freely available.

RELEASE HISTORY

Oct 16, 2026 · current package (UTC)

First ACD Drugs@FDA release. Eight related reference/accounting CSV tables, dictionaries, source/processing records, checksums and a checked Python example. No previous ACD edition is being compared.

Initial release: recorded submissions and source quality

Processing corrections and improvements for your purchased source edition are free. Genuinely new source editions may be sold separately. See the update policy and access instructions.

Check the downloaded CSV’s integrity

The metadata file inside the ZIP records this SHA-256 checksum for submissions.csv. A matching checksum confirms identical CSV bytes.

cfb5710f9410996576e8a57ef2b32e12c8c5cfacf5ca4ec1797fa783cd9c7409

READY TO PUT IT TO WORK?