What Is a Good TRIR? Construction Benchmarks, GC Cutoffs, and How to Lower Yours (2026)

Short answer: in construction, a TRIR under 3.0 gets you through most general contractor prequalification screens, a TRIR below the industry average of roughly 2.5 is genuinely good, and under 1.0 is elite — the range the safest large contractors live in. Zero is realistic for small crews and a legitimate multi-year goal for anyone.

Now the longer answer, because TRIR has become the safety credit score of construction — it decides which bid lists you're on, what your insurance costs, and how owners read your company before they've met you. Here's what the number means, where the cutoffs actually sit, and what actually moves it. (Want yours right now? The free TRIR calculator benchmarks your rate and DART in ten seconds.)

What TRIR Actually Measures

TRIR = (OSHA recordable incidents × 200,000) ÷ total hours worked.

The 200,000 is 100 full-time workers × 2,000 hours a year, so TRIR reads as "recordables per 100 workers per year," which lets a 12-person electrical shop be compared to a 4,000-person GC on the same scale.

A recordable is anything that lands on your OSHA 300 log: medical treatment beyond first aid, restricted work or job transfer, days away, loss of consciousness, or death. A bandage and back to work is first aid, not a recordable. Stitches, a prescription, or a lost shift is a recordable — and the difference between those two outcomes is often one guardrail.

DART is the same formula counting only the serious subset — cases with Days Away, Restricted duty, or Transfer. Prequal forms usually ask for both, which is why the calculator returns both.

The Benchmarks That Matter in 2026

• BLS construction average (NAICS 23): ≈ 2.5 recordables per 100 workers. Specialty trades (roofing, framing, drywall) tend to run higher; heavy civil and industrial work lower.
• The 3.0 line: the most common GC and owner prequalification cutoff. Above it, expect to be flagged, asked for a corrective action plan, or quietly left off the bid list.
• 1.0 and under: where safety-award-winning contractors and most owner-direct industrial work sits. Some petrochemical and data-center owners screen at 1.0 or better, full stop.
• Prequal platforms (ISNetworld, Avetta, Veriforce) grade you on TRIR, DART, EMR, and your written programs together — a strong TRIR with no documented safety program still scores as a red flag.

One number to pair with it: your EMR (experience modification rate). TRIR is incidents; EMR is what those incidents cost insurers. A TRIR spike this year becomes an EMR problem for the next three — that's the true cost of a recordable compounding.

The Small-Company Math Trap

Here's what prequal reviewers know and small contractors sometimes don't: at a small company, one incident detonates the rate.

• 10 employees ≈ 20,000 hours/year → one recordable = TRIR 10.0
• 25 employees ≈ 50,000 hours → one recordable = TRIR 4.0
• 100 employees ≈ 200,000 hours → one recordable = TRIR 1.0

So if you're a 15-person crew with one stitches-and-back-to-work incident, your TRIR looks catastrophic next to a 500-person GC's — while your actual safety culture might be better. Two defenses:

1. Report a 3-year average. Most prequal forms allow it, and it smooths the lumpy math. Keep your 300A summaries from prior years handy.
2. Attach context. A one-page narrative — what happened, what changed after — turns a scary number back into a story a reviewer can accept.

The flip side is just as true: at a small company, preventing a single incident is worth multiple points of TRIR. No other safety investment has that leverage.

What Actually Lowers TRIR

TRIR is a lagging indicator — by the time it moves, someone's already hurt. You lower it by working the leading indicators:

• Find hazards before they find people. A daily walk with fresh eyes beats a monthly audit. This is exactly what a 30-second phone scan is for — point it at the work area, get the hazards an inspector would flag, fix them the same morning.
• Job hazard analyses on the tasks that generate your recordables — which, industry-wide, cluster around the Fatal Four plus sprains and lacerations.
• Near-miss reporting. Every recordable was a near-miss first. Crews that log near-misses without blame catch the pattern one incident early.
• Weekly toolbox talks matched to the actual site conditions that week — not the same generic sheet every Monday.
• Get the log right. Over-recording first-aid cases inflates TRIR for nothing; under-recording real cases is an OSHA citation and torches your credibility on audit (what that costs).

The Bottom Line

TRIR	What it says
0	Achievable, especially under ~50 employees — and worth advertising
Under 1.0	Elite — opens owner-direct and industrial work
1.0–2.5	Better than the construction average — a selling point
2.5–3.0	Passes most screens, wins no points
Over 3.0	Expect flags, CAP requests, and lost bids

Run your own numbers in the TRIR calculator, and if the answer stings, remember the leverage: at most companies the whole rate hinges on one or two incidents a year. Catch one hazard before it becomes a recordable and you've moved the number that decides your bid list. That's the entire pitch for Site Safety AI — a 30-second scan every morning, every hazard flagged with a fix, and a timestamped record that you were looking.