How the calculators are checked

The primary sources, rounding rules and tests behind every TRIR Calc calculator, and the recordkeeping rules the tools don't model.

Updated September 19, 2026

On this page

Each calculator on TRIR Calc starts from a primary source, and each rule it applies can be traced to a paragraph or a form. This page lists those sources, then covers how cases are counted, how results are rounded, how the tools are tested, and what they leave out.

Primary sources

29 CFR Part 1904, OSHA's recordkeeping regulation, as published on eCFR. The copy the site works from is dated September 17, 2026.
OSHA's recordkeeping forms package, Forms 300, 300A and 301 with their instructions, which carry the incidence-rate instructions and the 300A worksheet for employees and hours.
BLS Appendix C, the Bureau of Labor Statistics guide to working out a firm's own incidence rate, used for its ABC Company example.
BLS news release USDL-26-0101, published January 22, 2026, used for the 2024 private-industry rate and its note on converting between rate bases.

Wording that a tool page quotes comes from copies of these pages saved when the site was built. Some rules the pages refer to, such as 1904.5 and 1904.6, have no saved copy, so those are linked and described rather than quoted.

How cases are counted

The rate calculators ask for each of the 300A's case columns and count total recordable cases as columns G + H + I + J added together: every line entry on the Form 300, deaths included.

OSHA's forms package offers two routes to the number. One is to count the line entries on the Form 300. The other is a shortcut on the 300A that adds up only columns H, I and J. Column G, deaths, sits outside that shortcut, so the two totals differ only when the log has a death on it. The calculators follow the line count.

Arithmetic and rounding

Rates. The TRIR, DART rate and days-away case rate each multiply a case count by 200,000 and divide by the hours worked. The exact formulas, with the columns each one counts, are on the TRIR calculator and the DART rate calculator. Since every input is a whole number, a rate is an exact fraction, and it's rounded only once, half up, to 2 decimals.
Two decimals is this site's choice. BLS prints its rates with 1 decimal, so round a result here to one place before you set it beside a BLS figure.
A different base. The case-characteristics part of USDL-26-0101 gives its days-away and job transfer or restriction rates per 10,000 full-time equivalent workers, not per 100. Move the decimal point two places left first; the DART page walks through it.
Full-time equivalents are the hours divided by 2,000, shown to 1 decimal and rounded half up.
The 300A's workforce figures. The average number of employees and the estimated hours are both rounded up to the next whole number, which is what the 300A worksheet instructs. A result that's already whole stays as it is.

Tests

Automated tests hold the tools to their sources:

Golden cases are worked by hand from the source and fix the expected output. They include BLS's ABC Company, where the site shows 16.92 and 7.69 against BLS's printed 16.9 and 7.7, and OSHA's Acme Construction example, where 830 ÷ 26 = 31.92 is rounded up to 32.
The recordability checker is run through every combination of the six general recording criteria. Each of the fourteen first-aid items is tested on its own, and again alongside a treatment that isn't on the list.
Every quotation on the site is compared with the saved source text. The words must match; line wrapping, spacing, dashes and the style of quote mark are ignored.

Keeping it current

The sources were last checked on September 19, 2026. They're rechecked every quarter, next in December 2026, and also whenever OSHA amends Part 1904. Any change that alters a result is logged on What changed.

What the tools don't model

Experience modification rate (EMR). An EMR is built from insurance data that insurers and rating bureaus hold, not from anything on the OSHA forms, so the site doesn't calculate one.
BLS industry comparisons. BLS publishes rates by industry, but none appear here until the figures have been checked against BLS's own table. The TRIR page gives the one benchmark in use, the 2024 private-industry rate.
The specific recording criteria in 29 CFR 1904.8 to 1904.12. The recordable checker applies only the general criteria of 1904.7.
Work-relatedness and new cases, which 1904.5 and 1904.6 govern. The checker asks you both questions and takes your answers as given.
State Plans. The tools follow the federal rule in Part 1904 and don't model any State Plan's differences.