How to calculate TRIR and DART

The formulas

Both cover a single calendar year for a single establishment, and both use the same denominator. DART is always less than or equal to TRIR, because every DART case is also a recordable case. If your DART comes out higher than your TRIR, something is wrong with the case counting.

If you just want the numbers, the TRIR and DART rate calculator does this arithmetic in your browser. The rest of this guide is about getting the inputs right.

Where 200,000 comes from

It is not arbitrary, and it is not a unit conversion. It represents the hours worked by 100 full-time equivalent employees over one year: 100 workers × 40 hours per week × 50 weeks.

The consequence is that TRIR is expressible in plain language: the number of recordable cases you would expect per 100 full-time workers per year. A TRIR of 3.0 means roughly three recordable cases per hundred full-time employees annually. That framing is worth using with executives, because "3.0" on its own invites the question "out of what?" — and the answer is what makes the number comparable between a site with 40 people and one with 900.

What counts in the TRIR numerator

Every case recorded on your OSHA 300 log for the period — no more, no less. If it earned a line on the log, it counts once, regardless of how serious it was. A recordable case involving a single day of restricted duty counts exactly as much toward TRIR as a fatality.

What does not count:

• First aid only cases, as OSHA defines first aid. These are not recordable, so they are not in the numerator.
• Near misses. No injury occurred, so there is nothing to record. This is why encouraging near-miss reporting cannot raise your TRIR.
• Non-work-related injuries, which never reach the log.

Count cases, not people and not days. One employee with three separate recordable injuries in a year contributes three cases.

What counts as a DART case

A DART case is a recordable case whose outcome involved at least one of:

• Days away from work — the employee could not work at all for one or more calendar days after the day of injury.
• Restricted work — the employee worked, but not their full normal duties.
• Job transfer — the employee was moved to a different job because of the injury.

Each qualifying case is counted once, even if it involved both days away and a subsequent period of restricted duty. DART counts cases, not days — a frequent source of inflated rates. A case classified as "other recordable" is in your TRIR but not your DART.

The relationship between the two rates is informative. A TRIR of 4.0 with a DART of 3.6 describes a workplace where nearly every recorded injury was serious enough to keep someone off their normal job. A TRIR of 4.0 with a DART of 0.8 describes something quite different, even though the headline number is the same.

Getting hours worked right

The denominator is where most miscalculations happen, and it is worth more care than the numerator because an error here shifts every rate you report.

• Use actual hours worked, from payroll where possible — not scheduled hours and not an assumed 2,080 per head.
• Include overtime. Overtime hours are hours of exposure and belong in the denominator.
• Exclude paid non-work time — holiday, vacation and sick leave are not hours worked.
• Include everyone the establishment's log covers, which generally means temporary and supervised staff whose day-to-day work you direct. If their injuries would be recorded on your log, their hours belong in your denominator. Leaving the hours out while keeping the cases in inflates your rate.
• Match the boundary to the numerator. Per-establishment cases require per-establishment hours. A company-wide denominator with one site's cases produces a meaningless figure.

A worked example

The following figures are hypothetical and are used only to show the arithmetic.

A single establishment, one calendar year. 180 employees, 372,000 actual hours worked including overtime. The 300 log shows nine recordable cases: one with days away, three with restricted duty, and five classified as other recordable.

Read together: about 4.8 recordable cases per 100 full-time workers, of which roughly 2.2 per 100 were serious enough to affect the person's ability to do their normal job. Rates are conventionally reported to two decimal places.

Severity rate

TRIR and DART both count cases and ignore how bad they were. Severity rate addresses that by looking at days rather than cases — commonly total lost and restricted days divided by the number of recordable cases, giving an average days-lost per case.

Be careful comparing severity rates between organizations: unlike TRIR and DART, there is no single universally applied formula, and some organizations normalise it against 200,000 hours as well. Check what someone means before comparing. Within one organization, tracked consistently, it is a useful counterweight to a TRIR that treats a fatality and a sprain identically.

What these rates do not tell you

This is the part most often skipped, and it matters more than the arithmetic.

• A low rate is not proof of a safe workplace. It is consistent with a safe workplace, and equally consistent with under-reporting, a lucky year, or a workforce that has learned reporting is unwelcome. Rates measure recorded outcomes, not risk.
• They are lagging indicators. They describe harm that has already happened. Near-miss reports, hazard reports, inspection findings and corrective-action closure rates tell you about the system while there is still time to change it.
• Small numbers are volatile. At a 60-person site, one additional case can move TRIR by more than a point. Treating that swing as a trend leads to reacting to noise. Below a few hundred thousand hours, look at multi-year patterns rather than year-on-year changes.
• They are blind to severity. A year with one fatality can produce a lower TRIR than a year of minor recordables.
• Targets distort them. A rate tied to bonuses is a rate under pressure, and the cheapest way to improve it is to record fewer cases. If you set targets, set them on leading activity — inspections completed, actions closed on time, near misses reported — rather than on the injury rate itself.

Use TRIR and DART for what they are good at: a normalised, comparable measure of recorded injury outcomes over time and between sites. Do not ask them to answer whether your program is working.