Three rates come off the same OSHA 300 log, and mixing them up is how safety numbers get quoted wrong. TRIR counts every recordable case. DART counts the subset that cost someone days away, restricted duty, or a job transfer. LTIR counts only days-away cases. Same formula, same 200,000-hour base, three different severities. Here's how to compute, read, and defend each one.
Total recordable incident rate, sometimes called TCIR, takes the year's recordable cases from column G of your OSHA 300 log, divides by the total hours all employees actually worked, and scales to a 100-worker-year base:
TRIR = recordable cases × 200,000 ÷ hours worked
Worked example: a 125-worker plant (250,000 hours) logs 5 recordables. TRIR = 5 × 200,000 ÷ 250,000 = 4.0. Two of those cases involved days away or restriction, so DART = 2 × 200,000 ÷ 250,000 = 1.6; one case kept a worker off the job entirely, so LTIR = 1 × 200,000 ÷ 250,000 = 0.8. Use actual payroll hours, overtime included. Substituting headcount × 2,000 quietly flatters overtime-heavy sites.
It's the annual hours of 100 full-time workers at 40 hours a week for 50 weeks: 100 × 40 × 50 = 200,000. Every rate therefore reads as "cases per 100 full-time-equivalent workers per year," which puts a 40-person shop and a 4,000-person plant on the same axis. It also explains why small sites lurch: at 60,000 hours, one recordable case is worth 3.33 points, and two cases land you at 6.67. A small-site TRIR moving between 3 and 7 usually signals arithmetic, not a collapsing safety culture.
Under 29 CFR 1904.7, a work-related injury or illness is recordable when it results in any of:
First aid alone — OSHA publishes the exact list — never makes the log. Prescription-strength medication, sutures, and immobilization are beyond first aid and do. The recordable decision follows the regulation's logic tree, not gut feel, and the log stays open for 5 years of updates.
| Industry sector | TRIR (TRC) | DART | Days-away rate |
|---|---|---|---|
| All private industry | 2.3 | 1.4 | 0.8 |
| Manufacturing | 2.7 | 1.7 | 0.8 |
| Construction | 2.2 | 1.3 | 0.9 |
| Transportation & warehousing | 4.4 | 3.5 | 2.1 |
| Retail trade | 3.0 | 1.8 | 1.0 |
| Health care & social assistance | 3.4 | 1.7 | 1.1 |
| Accommodation & food services | 2.6 | 1.2 | 0.8 |
BLS Survey of Occupational Injuries and Illnesses, 2024 rates (released January 2026). Private industry's 2.3 was the lowest in the series back to 2003. Benchmarking against the wrong sector is the classic error: a 2.5 TRIR is below-average for warehousing (4.4 sector average, 4.8 in warehousing proper) and above-average for construction (2.2).
Two separate obligations. The Form 300A annual summary must be certified by a company executive and physically posted at every establishment from February 1 through April 30. Separately, covered establishments submit data electronically through OSHA's Injury Tracking Application by March 2: 250+ employee establishments and 20-249 employee establishments in designated industries submit the 300A, and establishments with 100+ employees in designated high-hazard industries also submit Forms 300 and 301. Miss the window and you still have to submit; late is penalized, never is worse.
Three ways. OSHA's Site-Specific Targeting program pulls electronically submitted rates into inspection lists. The submitted data is publicly searchable, so journalists, unions, and sales prospects can look you up. And general contractors increasingly screen subcontractor TRIR during prequalification, typically wanting to see 1.0 or lower, which is why the metric earns its place on dashboard walls. An honest rate with a falling trend beats a massaged rate with a cliff in it.
Hours worked plus your 300 log counts, benchmarked against 2024 BLS sector averages, with the cases-allowed-at-benchmark number safety committees plan around.
OSHA Incident Rate Calculator →TRIR is breadth, DART is severity, LTIR is the worst outcomes, and all three divide by the same 200,000 hours. Compute them from actual payroll hours, benchmark against your own NAICS sector, and read small-site numbers as ranges rather than points. The companion math lives in our workers comp premium calculator when cases turn into claims, and in downtime cost when the same incidents stop production.
Many GCs and owners screen for a TRIR at or below 1.0 on prequalification forms, and anything under the 2.3 private-industry average (BLS 2024) is defensible. What matters more than the raw score is the trend and the denominator: reviewers discount rates from tiny sites because one case moves them several points. Keep three years of data and your hours-worked basis ready to explain.
No. OSHA publishes a specific first-aid list (bandages, cold packs, OTC meds at nonprescription strength, and so on), and cases receiving only those treatments are not recordable. The line is medical treatment beyond first aid: prescription-strength medication, sutures, immobilization, physical therapy. When in doubt, the 1904.7 decision logic, not intuition, decides.
They overlap but don't match. OSHA recordability is federal, log-based, and includes cases with no wage loss; workers' comp compensability varies by state, follows insurance rules, and includes wage-replacement cases that may or may not be OSHA-recordable. A recordable case with no lost time can generate zero comp claim dollars, and a commuting accident can be a comp claim without being recordable.