OSHA ITA API vs CSV Upload: Which Submission Method Fits Your Team?

CSV upload suits most employers with a few establishments; the ITA API pays off for many establishments filed each year from validated data. Both end in the same place: a submission in the ITA and a confirmation to keep.

The best ITA method depends on volume, repeatability, and validation controls, not technical pride.

Covered establishments submit injury data to OSHA's Injury Tracking Application (ITA) by March 2 each year for the previous calendar year; for 2026 data that is March 2, 2027 (29 CFR 1904.41). Three groups must submit: establishments with 250 or more employees that must keep records (300A data); establishments with 20 to 249 employees in an industry listed in Appendix A to Subpart E (300A data); and establishments with 100 or more employees in an industry listed in Appendix B (300A data plus 300 and 301 case data). Headcount means every individual employed at any time in the prior year, including part-time, seasonal, and temporary workers, and every submission includes the establishment's EIN.

CSV upload works well for many teams, while API submission can fit automated recordkeeping systems. Most ITA problems are establishment data problems: an old NAICS code, a parent-company EIN, or a headcount measured at the wrong level. The fix is to decide coverage per establishment in the fall, clean the EIN, NAICS, and address data then, and leave only validation and submission for February.

Whatever you use should make these easy to keep and retrieve:

• The coverage decision for each establishment: peak headcount, NAICS code, and which appendix applies
• The EIN used by each establishment
• The submitted file or payload and the ITA confirmation
• Notes on any corrections made after submission

1. Count establishments and filings per year.
2. Assess data quality at the source.
3. Decide who will check and submit.
4. Choose CSV or API, and test before February.

Technical teams should size the solution to the data: automation only helps once the source records are stable.

Keep these in mind when you compare approaches:

• Teams sometimes build API work too early when their source data is still unstable.
• High-volume multi-establishment employers benefit most from automated validation and repeatable exports.
• For 300 and 301 case data, OSHA excludes employee names and addresses and health care provider details from the submission (1904.41(b)(9)).

Spreadsheets fail quietly: a sort that renumbers rows, a tab copied for the new year, a formula that misses a new column. None of that shows on the printed form, which is why any system, including a spreadsheet, should be tested on real historical cases before it is trusted.

A company with 4 establishments uploads a CSV once a year in about an hour. A company with 180 establishments, whose incident system already holds validated establishment data, moves to the API: submissions are generated per establishment, responses are logged, and rejected establishments come back as a work list.

300log validates EIN and NAICS formats, produces ITA-ready CSV files for the 300A and for case data, and can authorize auto-filing of 300A data once you connect an ITA API token.

For a quick answer without an account, use the free ITA coverage checker.

• Correcting an OSHA ITA Submission After You Find an Error
• OSHA ITA FAQs: Submission Dates, Coverage, and Corrections for 2026 Data
• OSHA ITA Establishment ID FAQ: What Safety Teams Need Before Upload
• OSHA ITA 2027 Planning: Build the Filing Calendar Before January