What Is TRIR? How to Calculate Your Total Recordable Incident Rate

TRIR — Total Recordable Incident Rate — is the single number that procurement teams, insurers, and regulators ask for first. But calculating it correctly requires more than plugging numbers into a formula. You need to know what OSHA counts as recordable, which hours to include, and how to interpret the result.

TRIR = (Number of OSHA-recordable cases × 200,000) ÷ Total hours worked by all employees. The 200,000 constant represents 100 full-time equivalent workers working 2,000 hours each per year. This normalises every worksite to the same scale.

• A recordable case is any work-related injury or illness requiring medical treatment beyond first aid, resulting in days away from work, restricted work, job transfer, loss of consciousness, or death (OSHA 29 CFR 1904.7).
• Hours worked include all employees — full-time, part-time, temporary, and contractor hours if you supervise them day-to-day.
• Do NOT include: commuting (unless in company vehicle), eating/drinking, or personal tasks unrelated to work.

Let us walk through an example. A manufacturing site with 250 employees working 500,000 hours in a year reports 4 recordable injuries. TRIR = (4 × 200,000) ÷ 500,000 = 1.6. This means 1.6 recordable injuries per 100 FTE per year. The industry average for manufacturing is around 3.3 (BLS 2024), so 1.6 is above average.

If you are comparing TRIR across companies, make sure they use the same definition of recordable. Some global firms use their own stricter internal criteria, inflating their TRIR relative to OSHA-only reporting. Always ask: calculated per OSHA or per internal policy?

The fastest way to get clean TRIR data is to capture incidents in a system that classifies OSHA recordability at the point of entry — before the incident even reaches the safety manager. See the TRIR glossary entry for the full regulatory context and related metrics like LTIR and DART.

The work-relatedness decision is the one that decides whether a case enters the numerator, and it is the one the recordkeeping standard spends the most words on. Under 29 CFR 1904.5, an injury or illness is work-related if an event or exposure in the work environment either caused or contributed to the condition, or aggravated a pre-existing condition. The standard then carves out a set of exceptions in 1904.5(b) — the common cold, the flu, a case that arises from a voluntary wellness program, a case that is solely the result of a non-work event — and the exceptions are the ones a recordkeeper has to apply carefully, because the default is recordable. A TRIR calculated from a numerator that skipped the work-relatedness decision on the easy cases is a TRIR that quietly excludes the cases it should include.

The hours denominator is the one that is more often wrong than the numerator, and it is the one nobody checks until the rate moves for no reason. The denominator is total hours worked by all employees — full-time, part-time, and temporary — and it includes contractor hours only when the host supervises the contractor crew day-to-day. Hours paid but not worked (vacation, sick, holiday) are excluded; hours worked includes overtime. The common error is to use headcount times 2080 instead of actual hours, which overstates the denominator for a part-time or shift-heavy workforce and understates TRIR. The correct denominator is pulled from payroll, not estimated from a staffing plan, and the platform that classifies the case at entry also pulls the hours from the payroll feed so the two never drift.

The 200,000 constant is the normaliser, and it is worth knowing what it is so the rate reads correctly. 200,000 hours is 100 full-time-equivalent workers each working 2,000 hours in a year, so a TRIR of 1.6 means 1.6 recordable cases per 100 full-time-equivalent workers per year. The constant lets a 50-person shop and a 5,000-person plant be compared on the same scale, which is the whole point of the metric. A site that reports an absolute count of injuries without the rate is reporting a number that cannot be compared to anything, and the rate — not the count — is what procurement and insurers ask for.

The small-numbers problem is the one that makes TRIR volatile for small sites, and it is the one that small-site safety leaders have to explain every quarter. A site of 50 FTE working 100,000 hours a year that has zero recordables has a TRIR of 0.0; the same site with one recordable has a TRIR of 2.0. One event moves the rate by two points, and the year-over-year swing is noise, not trend. The defence is to report a three-year rolling TRIR for small sites and a one-year TRIR only for sites large enough that the rate is stable — and to never let a single quarter of a small site drive a procurement or insurance decision.

The DART and LTIR distinction is the one a recordkeeper has to hold straight, because the three rates measure different things and a regulator asks for all three. TRIR counts every recordable case. DART (Days Away, Restricted, or Transferred) counts the subset of recordables that involved days away from work, restricted duty, or job transfer — the cases that actually cost production. LTIR (Lost Time Injury Rate) counts the narrower subset with days away from work only. A site can have a healthy TRIR and an unhealthy DART if the recordables are concentrated in the days-away cases, and the DART rate is the one that signals severity where TRIR signals frequency. The three rates together — frequency, severity, lost-time — are the picture a board should see, not TRIR alone.

1. Classify recordability at the point of incident entry against the 29 CFR 1904.7 first-aid list, so the numerator is the count of cases the system already marked recordable.
2. Pull the hours denominator from the payroll feed, not from a staffing estimate, and include part-time, temporary, and day-to-day-supervised contractor hours.
3. Compute TRIR, DART, and LTIR from the same numerator split, so the three rates are consistent and the severity signal is not lost in the frequency number.
4. Report a three-year rolling TRIR for sites under 200 FTE and a one-year TRIR for larger sites, so the small-numbers volatility does not drive decisions.
5. Benchmark against the BLS industry average for the NAICS code, not against a generic average, and publish the benchmark alongside the rate so the number reads in context.

The procurement threshold is the use case that makes TRIR matter outside the safety team, and it is the one that pulls the metric into the contract. A typical construction or oil-and-gas pre-qualification asks for a three-year TRIR below the industry average and an EMR below 1.0, and a site that cannot produce the rate from a live query produces it from a spreadsheet that may or may not match the 300 log. The platform that holds the 300 log and the payroll hours produces the three-year TRIR on demand, and the on-demand number is the one a procurement team can submit without a week of reconstruction. For the full regulatory context and the related metrics, see the TRIR glossary entry and the LTIR and DART glossary entries; for the recordkeeping procedure, the OSHA 300 fastest-path post.

The multi-establishment question is the one that decides whether a company TRIR is a real number or an average of averages. Each establishment keeps its own 300 log and its own hours, and a company TRIR is the sum of recordables across all establishments divided by the sum of hours across all establishments — not the average of the establishment TRIRs. The two are the same only when every establishment works the same hours per person, which they do not. A company that reports the average of establishment rates is a company that lets a small establishment with one recordable swing the company number, and the procurement team that reads the company TRIR is reading a number that was not computed against the recordkeeping standard.