What counts as a DART case?

A DART case is a recordable injury or illness marked in column H (days away from work) or column I (job restriction or transfer) on the OSHA 300 log. Each qualifying case counts once, regardless of how many days it caused. Medical-treatment-only cases in column J are recordable but do not count toward DART.

What is the difference between DART and TRIR?

The math is identical; the scope is not. TRIR counts every recordable case, including medical-treatment-only cases. DART counts only the cases that resulted in days away, restriction, or transfer. Because every DART case is a recordable case but not the reverse, DART is always equal to or lower than TRIR.

How do I calculate total hours worked for DART?

Use actual hours on the clock from payroll, including overtime and excluding paid leave such as vacation and sick time. If you cannot pull exact hours, OSHA's convention is to estimate 2,000 hours per full-time employee per year and adjust for part-time and overtime. A bad hours denominator is the most common source of a wrong DART rate.

Can the DART rate be higher than TRIR?

No. DART cases are a subset of all recordable cases, so a correctly calculated DART rate is always equal to or lower than the TRIR for the same period. If your DART comes out higher, you have miscounted cases or used different hour totals for the two numbers.

Where does the 200,000 in the formula come from?

It is the number of hours 100 full-time employees work in a year: 100 workers x 40 hours per week x 50 weeks. OSHA and BLS use it as the common base for every incidence rate, so a DART or TRIR of 2.0 means the same thing at a 40-person plant and a 4,000-person plant.

How to Calculate DART Rate

Q: What is the DART rate formula?

DART rate = (number of DART cases x 200,000) / total hours worked by all employees during the year. A DART case is any OSHA-recordable case that resulted in days away from work, restricted duty, or a job transfer. The 200,000 represents 100 full-time employees working 40 hours a week for 50 weeks, which makes the rate comparable across employers of any size.

Q: What is a good DART rate?

There is no universal good number. Recent BLS private-industry DART has run near 1.5 per 100 full-time workers, but higher-hazard industries run well above that, so compare to your own NAICS industry code. The number that matters most is your own trend over time, read alongside leading indicators like near-miss reports and closed corrective actions.

The formula, a worked plant example, and how DART differs from TRIR by counting only the injuries that cost real days.

The DART rate is the number of workplace injuries and illnesses per 100 full-time workers that result in days away, restricted work, or job transfer, calculated as (DART cases × 200,000) ÷ total hours worked. It measures the more serious end of your recordable injuries.

TRIR counts every recordable injury; DART counts only the ones that actually kept someone off the job or on light duty. That is what makes DART the number safety leaders and insurers watch: it strips out the minor recordables and leaves the injuries that cost real days. This post covers what DART counts, the formula, a worked calculation, how it differs from TRIR, and how to read it against the national benchmark. It is educational, not legal advice.

What does the DART rate measure?

DART stands for Days Away, Restricted, or Transferred, and it counts the OSHA-recordable cases that reach that severity. A case counts toward DART if the injury or illness resulted in the employee taking days away from work, being placed on restricted or light duty, or being transferred to a different job. Each qualifying case counts once, no matter whether it caused one restricted day or forty.

Those case types come straight off the OSHA 300 log. Column H marks cases with days away from work; column I marks cases with job transfer or restriction. A case checked in either column is a DART case. Column J, other recordable cases that needed medical treatment beyond first aid but no days away or restriction, is recordable but is not a DART case. That single distinction is the whole idea: DART is the log minus the medical-treatment-only cases.

Every DART case is a recordable case, but not every recordable case is a DART case. The difference is the medical-treatment-only column.

What is the DART rate formula?

The formula is (number of DART cases × 200,000) ÷ total hours worked by all employees during the year. The 200,000 is the anchor that makes the rate comparable across companies of any size: it is 100 full-time employees working 40 hours a week for 50 weeks. So a DART rate of 2.0 means two DART cases per 100 full-time workers per year, whether you employ 40 people or 4,000.

The same 200,000-hour base OSHA and BLS use for every incidence rate, so results compare directly across employers.

Total hours worked means actual hours on the clock, overtime included, paid leave excluded. Do not estimate it from headcount if you can pull it from payroll; the hours denominator is where most bad rates come from. If you must estimate, OSHA's convention is 2,000 hours per full-time employee per year, then adjust for part-time and overtime.

How do you calculate a DART rate?

Walk one plant through it. Say a plant worked 450,000 hours last year and its 300 log shows 5 cases with a check in column H or column I.

Count the DART cases. Add up every case on the OSHA 300 log with a mark in column H (days away) or column I (restriction or transfer). Do not double-count a case that has both; it is one DART case. Here: 5.
Total the hours worked. Pull actual hours from payroll for every employee across the year, including overtime, excluding vacation and sick leave. Here: 450,000.
Multiply the cases by 200,000. 5 × 200,000 = 1,000,000.
Divide by total hours. 1,000,000 ÷ 450,000 = 2.22.
Round and record. Round to two decimals. This plant's DART rate is 2.22, meaning about 2.2 serious cases per 100 full-time workers that year.

Run the same plant's TRIR for contrast. If the log also shows 4 medical-treatment-only cases, total recordables are 9, and TRIR = (9 × 200,000) ÷ 450,000 = 4.0. Same plant, same year: TRIR 4.0, DART 2.22. DART is always equal to or lower than TRIR, because every DART case is a recordable case but not the reverse. If your DART ever comes out higher than your TRIR, you have a counting error.

How is DART different from TRIR?

Scope, not math. The formula is identical; the case count is not. TRIR counts every recordable case, including the medical-treatment-only ones that never cost a day. DART counts only the cases that put someone off work, on restriction, or in a different job. See our full guide to how to calculate TRIR for the wider metric.

Same denominator, different numerator. DART strips out the medical-treatment-only cases TRIR includes.

Why track both? TRIR tells you how often people get hurt at all; DART tells you how often they get hurt badly enough to lose time. A plant can drive its TRIR up on purpose by encouraging reporting of small injuries, which is healthy, while its DART holds steady or falls because the serious injuries are actually going down. Reading them together keeps you from mistaking better reporting for worse safety.

What is a good DART rate?

Lower than your industry's, and trending down. There is no universal "good" number, but the Bureau of Labor Statistics publishes the national benchmark every year, and the private-industry DART rate has hovered around 1.5 cases per 100 full-time workers. Higher-hazard sectors run well above that, so the honest comparison is to your own NAICS industry code, not to the all-industry average. BLS lets you look up the rate for your specific industry.

The number that matters most is your own trend. A DART of 2.2 that was 3.5 two years ago is a program working; a DART of 1.0 that has crept up three years running is a warning even though it beats the benchmark. Pair the lagging DART number with leading indicators, near-miss reports closed corrective actions, audit findings resolved, so you are steering by what is coming, not only by what already hurt someone. A regular workplace safety audit is where those leading indicators get generated.

Why does the DART rate matter beyond compliance?

Because people outside your safety office use it to make decisions about you. Workers' compensation insurers read DART alongside your experience modification rate to price your premium, so a rising DART can cost money long before it costs a regulator's attention. General contractors and large customers routinely require a DART rate in prequalification, and many set a ceiling: above it, you do not get to bid. Corporate boards put DART on the operations scorecard because it is a severity number that a non-safety executive can read at a glance. In short, DART travels. It shows up in insurance renewals, customer audits, and bid packages, which is why the number has to be defensible and the log behind it has to be clean.

It is also a fairer number to compare across sites than a raw injury count. A 50-person line and a 500-person plant cannot compare totals, but they can compare DART rates, because the 200,000-hour base normalizes for size. That is what makes DART useful for benchmarking one plant against another inside the same company and against the national figure for the same industry.

What mistakes make a DART rate wrong?

Most bad DART rates come from the inputs, not the formula. The recurring errors:

Guessing the hours. Estimating the denominator from headcount instead of pulling actual payroll hours is the single most common error, and it moves the rate in whichever direction the guess is off.
Missing restricted-duty cases. A worker kept on light duty with no days away is still a DART case. Plants that only track "lost time" undercount DART by leaving out every column-I case.
Counting first aid. A cut that got a bandage and nothing more is not recordable and not a DART case. Sweeping first-aid events into the count inflates the rate.
Double-counting. A case with both days away and later restriction is one DART case, not two.
Comparing to the wrong benchmark. Measuring a fabrication shop against the all-industry average makes a normal rate look alarming or a poor one look fine.

Every one of those traces back to how well the injury record is kept. If restricted-duty days and hours worked are reliable, the DART rate takes care of itself.

What do the numbers say?

The primary sources for the formula and the benchmark:

The 200,000-hour base and recordable-case definitions come from OSHA's recordkeeping rule, 29 CFR Part 1904 and the OSHA 300 log.
The Bureau of Labor Statistics reports the national DART rate and its components in its annual Survey of Occupational Injuries and Illnesses; recent private-industry DART has run near 1.5 per 100 full-time workers.
BLS lets you pull the DART rate for your own industry with its Injuries, Illnesses, and Fatalities data tools which is the only apples-to-apples benchmark.

The weak link in most DART calculations is not the arithmetic, it is the log. When injuries, restricted-duty days, and hours worked live in three different spreadsheets, the case count drifts and the denominator is a guess. Harmony is an AI-native layer that connects machines, software, and paperwork into one operational layer, with no rip-and-replace: incidents, restricted-duty tracking, and the records behind them become structured, searchable data instead of scattered files, so the number you report is the number that actually happened. That is the same backbone a broader EHS audit runs on. Harmony is not an OSHA-recordkeeping product, but it keeps the source data clean enough to trust the rate. See how it works in a plant like yours in the CLS case study.