Fatal Accident Rate (FAR) is the number of work-related fatalities per 100 million hours worked, a lagging measure used to compare fatality risk across organizations, industries and activities on a common exposure basis.
The fatal accident rate (FAR) is the number of work-related fatalities per 100 million hours worked by a group of people over a defined period. The figure of 100 million hours is roughly the working lifetime of 1,000 people, which is why the measure became popular in UK process safety and the oil and gas sector: a FAR of 1 means about one death per thousand working lifetimes.
Because the denominator is exposure hours, FAR lets organizations of very different sizes compare fatality risk on a common scale. It also shows up in quantitative risk assessment, where engineers estimate an individual's FAR for a specific role or facility and compare it against tolerability criteria. The term is used loosely, though. Some companies and regulators quote fatalities per 100,000 workers or per million hours instead, so always check the denominator before comparing two numbers.
FAR = (number of fatalities ÷ total hours worked) × 100,000,000
Hours worked should include all employees in scope, and companies should state clearly whether contractor hours and contractor fatalities are included. In oil and gas reporting, employer and contractor figures are often reported both separately and combined.
In the United States, the Bureau of Labor Statistics publishes fatal work injury rates through the Census of Fatal Occupational Injuries (CFOI). The BLS expresses these as fatalities per 100,000 full-time equivalent (FTE) workers, calculated from hours:
Rate = (N ÷ EH) × 200,000,000
Here N is the number of fatal injuries, EH is total hours worked by all employees in the group during the year, and 200,000,000 is the hours worked by 100,000 full-time workers at 40 hours a week for 50 weeks a year. For 2024, the BLS reported 5,070 fatal work injuries in the US and a rate of 3.3 per 100,000 FTE workers.
The BLS base of 200 million hours is exactly twice the FAR base of 100 million. So, for the same data, the BLS-style rate is double the FAR. A BLS rate of 3.3 per 100,000 FTE corresponds to a FAR of about 1.65. Getting this wrong is a common reason benchmark comparisons look better or worse than reality.
The figures below are hypothetical and for illustration only.
A drilling contractor runs several rigs with a combined workforce of about 2,400 people, including its own employees and subcontractors. Over five years, the company records 58 million hours worked and two fatalities: one from a dropped object on the rig floor and one in a vehicle rollover on a lease road.
Now look at a single year. In year three, the company worked 11.5 million hours and had one of those fatalities. The annual FAR for that year is 8.7. In each of the other four years it was zero. The five-year figure of 3.45 is a far more honest picture of risk than either the year-three spike or the zero years, which is why fatality rates are usually rolled over multiple years for organizations of this size.
The incident rate, or total recordable incident rate (TRIR), counts all OSHA recordable injuries and illnesses per 200,000 hours worked, about 100 full-time workers for a year. It captures far more events, from stitches to lost-time injuries, so it moves faster and is easier to trend. FAR counts only deaths.
A site can drive its TRIR down by reducing minor injuries while fatality risk stays flat, because the hazards that cause sprains and cuts are often not the ones that kill. Many organizations now track serious injury and fatality (SIF) potential separately for this reason. A low TRIR is not evidence of a low FAR.
The severity rate measures lost workdays per set number of hours. It reflects how serious injuries are, but some systems count fatalities as a fixed number of lost days, which makes it a blunt tool for comparing fatality risk.
Fatalities are rare at the level of a single company. A business working 2 million hours a year can go a decade without a death and then have two in one week. Its annual FAR swings from zero to a very large number on events that tell you little about whether risk changed. FAR is statistically meaningful mostly for large populations: whole industries, national workforces or very large employers over several years.
FAR measures outcomes after people have died. It cannot warn you in advance. Leading indicators, such as critical control verification rates, high-potential near miss reports and audit findings on life-saving rules, are more useful for steering day-to-day prevention.
Whether to include contractors, travel to and from remote sites, occupational disease deaths, or deaths occurring days after an incident differs between organizations and countries. Two FARs calculated on different rules are not comparable.
A company-wide FAR averages low-risk office hours with high-risk field hours. Calculating FAR by activity, such as driving, lifting or work at height, is more informative than one blended number.
The only acceptable target is zero, but as a benchmark, compare against your own industry using the same denominator and inclusion rules. Industry associations, such as those in oil and gas, publish multi-year fatal accident rates for their members. Comparing a single company's FAR against a national all-industry rate is rarely meaningful, because the mix of work is so different.
100 million hours approximates the combined working lifetimes of 1,000 people, each working about 2,000 hours a year for 50 years. That makes the number intuitive for risk discussions: a FAR of 2 means roughly two deaths across a thousand careers. It also produces whole-number values for most industries, where smaller bases would produce awkward decimals.
No. OSHA requires employers to report work-related fatalities to OSHA within eight hours and to record them on the OSHA 300 log where recordkeeping applies, but it does not require any fatality rate calculation. FAR is a voluntary metric used for internal performance tracking, client prequalification questionnaires, industry benchmarking and quantitative risk assessment.
For most purposes, yes. Contractors often perform the highest-risk work, so leaving them out makes the rate look artificially low. The common practice in oil and gas and mining is to report employee, contractor and combined rates, each with matching hours. If contractor hours are not tracked accurately, the combined rate is only as reliable as those estimates.
Part of SafetyIQ's EHS glossary: plain-English definitions of workplace health and safety terms.
> Keep learning
> See it in action
Capture hazards, run inspections, track training and close out corrective actions in one platform built for high-risk, field-heavy work.
Book my free demo
Get a demo built around your sites, your teams, and the way your people work.
EHS, journey, and fatigue management software for field-heavy teams.
.png){kind=logo}
.png)