You calculated your TRIR. Maybe it came out to 3.1, or 5.4, or 0 — and the number, by itself, tells you nothing. A 3.1 is excellent for a nursing home, unremarkable for a plumbing contractor, and alarming for a physician's office. The only way to read an incident rate is against the rate for employers who do the same kind of work you do — and the authoritative source for that comparison is the Bureau of Labor Statistics' Survey of Occupational Injuries and Illnesses (SOII), which publishes injury and illness incidence rates by industry every year.
This post is the reference for the 2024 data — the most recent available, released January 22, 2026 — organized the way a small employer actually uses it: the headline numbers, a benchmark table for the industries where small businesses live, and then the part most benchmark articles skip: what the comparison does and doesn't mean when your whole company is 15 people. (If you need the formulas themselves — what counts in the numerator, how to get the hours right — that's covered in our plain-English guide to TRIR and DART rates. This post assumes you have a rate and want to know what it means.)
The 2024 Headlines
From the BLS release (USDL-26-0101, January 22, 2026), covering calendar year 2024:
- Private industry employers reported 2.5 million nonfatal workplace injuries and illnesses — down 3.1 percent from 2023, and the lowest count in the data series going back to 2003.
- The total recordable case (TRC) rate was 2.3 cases per 100 full-time equivalent workers, down from 2.4 in 2023 — also the lowest on record for the series.
- Injuries alone occurred at 2.2 cases per 100 FTE. Illnesses fell to 13.9 cases per 10,000 FTE (from 19.0 in 2023), driven mostly by a 46 percent drop in respiratory illness cases — the continuing post-COVID normalization.
- No industry sector's rate increased in 2024. Five sectors decreased, including health care and social assistance (3.6 → 3.4) and information (1.0 → 0.7, the lowest of any major sector).
One vocabulary note before the table: BLS says "TRC rate" where most safety professionals say "TRIR." They are the same number — total OSHA-recordable cases × 200,000 ÷ hours worked. BLS also publishes a DART rate for most industries: cases involving days away from work, job transfer, or restriction, on the same per-100-FTE basis.
2024 Benchmarks for Common Small-Business Industries
All rates are per 100 full-time equivalent workers, private industry, all establishment sizes, from BLS SOII 2024. Find your industry — or the closest published level to your NAICS code — and read both columns: TRIR is how often people get hurt; DART is how often it's serious enough to cost work time.
| Industry (NAICS) | TRIR (TRC rate) | DART rate |
|---|---|---|
| Private industry overall | 2.3 | 1.4 |
| Construction of buildings (236) | 2.2 | 1.3 |
| Specialty trade contractors (238) | 2.3 | 1.4 |
| — Roofing contractors (238160) | 2.4 | 1.7 |
| — Electrical contractors (238210) | 1.8 | 1.1 |
| — Plumbing, heating & A/C contractors (238220) | 3.0 | 1.7 |
| Food manufacturing (311) | 3.3 | 2.3 |
| Fabricated metal product mfg. (332) | 3.2 | 1.8 |
| Furniture manufacturing (337) | 3.5 | 2.1 |
| Sawmills (321113) | 4.2 | 2.8 |
| Building material & garden supply dealers (444) | 3.3 | 2.3 |
| Grocery stores (4451) | 4.0 | 2.6 |
| Truck transportation (484) | 2.9 | 2.2 |
| Couriers and messengers (492) | 8.0 | 6.6 |
| Warehousing and storage (493) | 4.8 | 4.1 |
| Landscaping services (561730) | 3.0 | 2.0 |
| Offices of physicians (6211) | 1.8 | 0.4 |
| Hospitals (622) | 5.1 | 2.1 |
| Nursing & residential care facilities (623) | 5.5 | 3.6 |
| Food services and drinking places (722) | 2.4 | 1.0 |
| — Full-service restaurants (722511) | 2.2 | 0.8 |
| Automotive repair and maintenance (8111) | 1.9 | 1.1 |
A few patterns worth noticing. The industries the public assumes are most dangerous are not the ones at the top of the table: construction of buildings, at 2.2, sits below the all-industry average, while couriers (8.0), warehousing (4.8), and nursing and residential care (5.5) run two to three times the national rate. And the TRIR-to-DART ratio varies a lot by industry — in warehousing, roughly 85 percent of recordables cost work time; in full-service restaurants, closer to a third do. If your DART is close to your TRIR, your cases are running more serious than your industry's mix.
Where to find your own NAICS code
Your six-digit NAICS code is on your federal tax return (Schedule C or the business return), your workers' comp policy, or your D&B listing — or search your business description at census.gov/naics. BLS publishes 2024 rates on the 2017 NAICS vintage, and not every six-digit industry gets its own published rate: when yours isn't published, use the closest rollup (five-, four-, or three-digit prefix). Our free rate calculator does this resolution automatically — enter your NAICS code and it walks down the ladder to the deepest published benchmark and tells you which level it matched.
The Small-Employer Problem: One Case Moves Your Rate More Than Anything You Did
Here is the part that matters most below about 50 employees, and that benchmark articles written for corporate safety departments never mention.
The rate formula divides by hours worked and multiplies by 200,000 — the hours of 100 full-time workers. A 15-person shop puts in roughly 30,000 hours a year. Run the arithmetic:
- Zero recordable cases → TRIR 0.0. You beat every industry in the table.
- One recordable case → TRIR 6.7. You are now nearly three times the national average and worse than any industry above except couriers.
Nothing about your safety program changed between those two years — one employee needed stitches. At small hours denominators, TRIR is a coin-flip statistic year to year, and treating a single year's number as a verdict (in either direction) is innumerate. What a small employer can honestly do with the benchmark:
- Compare a multi-year average. Three years of cases over three years of hours smooths the coin flips into something closer to a signal.
- Watch DART separately. Whether cases cost work time is partly luck, but a DART that persistently tracks your TRIR means your incidents are serious, not just frequent — and the day counts that feed DART are calendar days with specific rules, so a counting error moves this rate too.
- Treat the trend as the metric. Down and to the right over three years beats any single-year comparison to the table.
A benchmark is context, not a compliance obligation
Nothing in 29 CFR Part 1904 requires you to have a TRIR below your industry average — or below anything. The recordkeeping rule requires accurate records, on time, kept and posted correctly. An employer with a TRIR of 6 and an accurate log is in compliance; an employer with a TRIR of 1 built on unrecorded cases is not, and under-recording is what inspectors are trained to look for. Never manage the number instead of the injuries.
Who Actually Reads Your Rate — and Which Data They Use
Three audiences compare your rates to these benchmarks, and it's worth knowing they don't all use the same data:
Insurers and general contractors use the number you give them, usually straight off your 300A. Many GC prequalification forms ask for three years of TRIR and DART ("EMR and OSHA rates"), and some apply hard cutoffs — commonly your industry average or a fixed threshold. This is where the benchmark touches revenue: for a subcontractor, a rate above the 238-series averages in the table can mean lost bids before anyone reads a safety program.
OSHA does not use SOII to pick inspection targets — SOII is an anonymized statistical survey. What OSHA uses is the establishment-level data you submit through the Injury Tracking Application every March 2, from which it computes your rates and feeds its Site-Specific Targeting program. Same underlying 300-log arithmetic, different pipeline — and it's your own submitted data, not an industry average, that puts an establishment on that list.
BLS itself collects SOII from a sample of employers each year (if you're sampled, participation is mandatory), and the survey rides directly on your OSHA 300 log — the survey form is essentially "transcribe your log." Accurate recordkeeping is what makes these benchmarks exist at all.
How Serious Are the Cases Behind the Averages?
The 2024 release also carries BLS's biennial case detail for 2023–2024, which is useful for calibrating expectations when you do have a case:
- Cases with days away from work made up 61.5 percent of all DART cases, with a median of 8 days away.
- Cases with job transfer or restriction only made up the rest, with a median of 15 days of restriction — restricted-duty cases run longer than days-away cases, which surprises people until they think about how return-to-work programs work.
- The largest driver of DART cases across the two years was overexertion, repetitive motion, and bodily-condition events (about 946,000 cases), ahead of contact-with-object incidents (about 860,000).
If an employee is out for two weeks after a lifting injury, that is not an outlier — it's the median neighborhood. What decides whether the case lands in your K column or your L column (and therefore in your DART numerator) is the outcome classification, which is exactly the place recordability and outcome calls have to be made correctly before any rate is worth benchmarking.
Common Questions
What is a "good" TRIR? There is no universal threshold. Below your industry's published rate is the defensible claim; below it on a three-year average is the credible one. For contract prequalification, the practical bar is whatever the GC's form says — commonly "at or below industry average."
Is a TRIR of 5 high? Against private industry overall (2.3), yes — more than double. Against nursing care (5.5) or hospitals (5.1), it's average. Industry first, then judgment.
Does OSHA fine you for a high TRIR? No. There is no citation for a high rate. Rates matter to OSHA as a targeting signal (via your ITA submission, not SOII), and recordkeeping violations are cited for inaccurate or late records — not for the injuries themselves being numerous.
Where do I get the full BLS table? bls.gov/iif — the "Nonfatal injuries and illnesses tables" page carries the complete industry table (Table 1) for every published NAICS level. The 2025 reference-year data should arrive on roughly the same schedule, in late 2026 or early 2027.
Benchmarking Without the Spreadsheet Step
The comparison only means something if the inputs are right: every recordable case captured (and nothing over-recorded), day counts kept to the calendar-day rules, and hours worked totaled from payroll rather than estimated. LogStead computes TRIR and DART from your 300 log and hours as you keep them, so the rate you benchmark is the rate your records actually support — and the free rate calculator runs the same 2024 BLS benchmark lookup shown in this post against any NAICS code, no account required.