To calculate lost time injury frequency rate, multiply the number of lost time injuries recorded in your reporting period by 1,000,000, then divide that number by the total hours worked over the same period by everyone in scope. Most companies can do the arithmetic in under ten minutes. The slow part is getting the numerator and the denominator to describe the same population over the same dates, which is where rates usually go wrong.
The three inputs are simple:
- Number of lost time injuries — work-related cases that stopped someone doing at least one full shift or a full day.
- Total hours worked — actual exposure hours by every person in your agreed scope, not scheduled or paid hours.
- The multiplier — normally 1,000,000, which scales your rate to a per-million-hours basis so a 20-person shop and a 20,000-person contractor can be compared.
Below is the full process, including the scope decisions that most employers have to argue about and the mistakes that quietly inflate a rate.
Table of Contents
What You Need Before You Calculate Anything
You need five things, and four of them are decisions rather than data. Getting them written down first saves the awkward conversation where a client compares your rate with theirs and the two numbers describe different populations.
- A written definition of a lost time injury. Fix the threshold in advance. The common international definition is any injury or illness that prevents a person working at least one full shift or one full day, which is the basis used by ILO-aligned reporting and by Safe Work Australia guidance. Some employers count a shift instead of a day for night or rotating crews, and that choice changes the numerator.
- The reporting period. Calendar month, calendar quarter, or a rolling twelve months. A rolling twelve months is the usual convention in contracts and collective agreements because it smooths out the swing you get from small numbers.
- An incident list for that period, with the date, the person, the classification (first aid, medical treatment, restricted duty, lost time, fatality), and the basis for that classification.
- A reliable hours total from payroll or the timekeeping system for the same period and the same population.
- A stated multiplier and scope note — for example, “LTIFR per 1,000,000 hours, employees plus supervised contractors, rolling twelve months to June.”
If the numbers come from an OSHA 300 log in the United States, the recordable and lost-day definitions come from 29 CFR 1904, and the same method produces the incident rate figures calculated in our guide to how to calculate the total recordable incident rate.
Step-by-Step: How to Calculate Lost Time Injury Frequency Rate
Step 1: Define the Reporting Period and Workforce
Write down the start and end date first, then decide who is in scope. A defensible scope normally includes full-time and part-time employees, supervisors and managers who work on site, and temporary or agency staff working under your control. It often includes contractor personnel when you supervise their hours directly, and excludes them when you only receive a scope-of-work summary.
Both choices are defensible. Mixing them is not. Multi-location employers have to decide whether the head office, warehouse and remote staff sit in the same denominator as the site crews, and they should say so in writing.
Check: add up headcount by month across the period and confirm nobody appears in two business units for the same hours. If a shared-services manager is split across two cost centres, apportion the hours once, not twice.
Step 2: Count Recordable Lost-Time Injuries
Count each qualifying case once, no matter how many body parts were hurt or how many shifts were lost. A worker who misses ten shifts after one injury is one lost time injury, not ten. That single rule removes most double counting.
Cases to exclude: first aid only (including ice packs, plasters and on-site nurse treatment), and any case that stays inside your medical-treatment-only threshold if you have set one. Restricted duty is the grey area that generates the most disagreement. Decide and document one rule: either count an injury when someone is off work entirely, or count it when they are formally restricted from their normal duties. Applying the first rule in one quarter and the second in the next makes your trend line meaningless.
Fatalities always count. So do cases carried over from a previous period if the injury occurred in a period you are reporting, using the date of the injury rather than the date it was closed out.
Check: reconcile your count against the source register line by line and write a one-line reason next to anything you excluded. A figure of six that someone can trace to six source records is defensible in an audit; a figure of six with no audit trail is not.
Step 3: Calculate Total Exposure Hours
Exposure hours are the hours people actually worked during the period, across the population defined in Step 1. Overtime counts at its actual value, so an eight-hour overtime shift contributes eight hours, not four. A double shift contributes sixteen.
Here is the include or exclude list that most disputes come down to:
| Category | Treatment | Note |
|---|---|---|
| Regular hours | Include | Actual hours clocked, not the roster |
| Overtime and double shifts | Include | At actual hours worked |
| Weekend and public holiday shifts | Include | Same as any other shift |
| Paid but not worked time | Exclude | Sick leave, holidays, paid leave |
| Contractor hours you supervise directly | Include, if in your stated scope | Needs the contractor hours on your invoice or timesheet |
| Subcontractor hours outside your control | Exclude | Report them in the contractor’s own figures instead |
| Travel time between sites | Usually exclude | Include only if you state that you do not |
| Standby or on-call at home | Exclude | No exposure occurred |
| Remote and home-office hours | Contestable | Include only where a real hazard exposure exists; state the choice |
| Shift handover briefings not on the clock | Include if paid | Consistency matters more than the choice |
For a straightforward workforce the total is employees multiplied by average hours per week multiplied by weeks, adjusted for overtime, leave and part-time patterns. Pull it from payroll rather than estimating it, because an estimate is the single most common reason a rate looks better than it should.
Check: total hours must cover every month in the period, including months when nobody was injured. A denominator built only from months with incidents produces a rate that is too high.
Step 4: Divide Injuries by Hours Worked
The core formula is:
LTIFR = (Number of lost time injuries × multiplier) ÷ total hours worked
Worked example, construction, rolling twelve months: 6 lost time injuries, 2,500,000 total hours worked.
6 × 1,000,000 = 6,000,000
6,000,000 ÷ 2,500,000 = 2.4
Read that as 2.4 lost time injuries per million hours worked. A smaller employer with 2 injuries over 400,000 hours gets (2 × 1,000,000) ÷ 400,000 = 5.0, so the smaller company is doing worse on the rate even though it had fewer injuries in absolute terms. That comparison only works because both used the same multiplier, definition and population.
In a spreadsheet the calculation is one cell. If B2 holds the injury count and B3 holds the hours: =(B2*1000000)/B3. For the 200,000 variant, change the number to 200000. Format the cell to two decimals.
Step 5: Express and Check the Result
Report the number, then the context, in this order: rate to two decimal places; injuries per million hours worked; the multiplier used; the injury definition used; the reporting period; the population included. A bare “2.4” in a board pack invites the wrong comparison.
Also report a 12-month rolling rate alongside any monthly figure. Monthly rates swing wildly on small numbers, and a month with zero injuries gives 0.00, which reads like perfection rather than an absence of events. Then check the four things that invalidate a calculation: numerator and denominator covering the same dates, the same population, a definition you can point to in writing, and a rounding rule applied consistently. If those four hold, the number is defensible.
How to read the trend matters more than the single value. A rate that falls from 4.1 to 2.2 over a year is a good sign; a rate that rises while hours worked fall is usually an hours issue, not a safety issue. For the analysis side of this work, how to analyze injury data for trends in seven steps walks through the segmentation.
1,000,000 or 200,000: Which Multiplier Applies to You
This is the most common question on the topic, and the answer depends on the framework you are reporting under rather than on anything you prefer.
| Multiplier | Typically used with | Injuries counted | Same data expressed as |
|---|---|---|---|
| 1,000,000 | ILO-aligned practice, ISO 45001 organizations, most Australian and European contracts, LTIFR as commonly used internationally | Lost time injuries (one full shift or day lost) | Baseline |
| 200,000 | US OSHA recordkeeping: incident rate, DART, TRIR | Recordable cases under 29 CFR 1904; DART excludes recordables with no lost or restricted day | Multiply the 1,000,000 rate by 0.2 |
| 200,000 | ANSI/AS 1885.1 style lost workday counting in some US contracts | Lost workday and restricted duty cases | Same conversion |
A rate of 2.4 per million is 0.48 per 200,000 hours. Neither number is wrong; publishing them side by side without naming the multiplier is.
Related rates use the same arithmetic with a different numerator: MTIFR counts medically treated injuries, FAIFR counts first aid cases, LTISR counts days lost per million hours worked, and RIFR counts restricted duty cases. ISO 45001 certification does not prescribe a single frequency formula, so organizations write their own definition into the management system and stick to it.
Common Mistakes That Distort the Rate
- Using scheduled or paid hours instead of exposure hours. Paid hours include annual leave and sick days nobody worked. Fix: take the hours figure from the timekeeping system.
- Counting one incident more than once. Multiple injuries in a single event, or a case re-opened after a second absence, inflate the count. Fix: one case per person per injury event.
- Including first aid cases. They belong in FAIFR, not the numerator of LTIFR. Fix: exclude anything treated as first aid only.
- Mismatched periods. A twelve-month numerator against a quarterly hours total produces nonsense. Fix: both figures from the same dates, every time.
- Mixing multipliers across periods or sites. A 0.48 followed by a 2.4 looks like a fivefold spike when it is just two conventions. Fix: convert everything to one basis before charting.
- Changing the LTI definition without restating history. Switching from shift to day mid-year breaks the trend. Fix: recalculate prior periods under the new rule or annotate the break.
- Counting restricted duty by one rule now and another later. Fix: pick one threshold, document it, apply it retrospectively.
- Letting the hours total follow the injury list. Adding hours only for months where an incident occurred inflates the rate. Fix: the denominator covers the whole period unconditionally.
Two further checks worth doing before the number goes out: confirm with HR that hires and leavers during the period are captured for their full employment, and confirm that anyone off on long-term disability is handled under your stated rule rather than silently dropped from the hours.
Frequently Asked Questions
Is LTIFR 1,000,000 or 200,000?
Both are correct in their own framework. The 1,000,000 multiplier is the international convention used for lost time injury frequency rate, and it matches ILO-aligned practice, ISO 45001 organizations and most Australian and European contracts. The 200,000 basis is the US OSHA convention used for the incident rate, DART and TRIR. Convert between them by multiplying the per-million rate by 0.2, and always state which one you used.
What counts as a lost time injury?
A work-related injury or illness that stops a person doing at least one full shift or a full day of their normal work. First aid only cases are excluded. Each person counts once per injury event, no matter how many shifts are missed or how many body parts are affected. Fatalities always count. Restricted duty cases are a judgement call: set one documented threshold, such as formally restricted from normal duties, and apply it to every period you report.
Do contractors and agency staff count in the hours worked figure?
It depends on your written scope, and the honest answer is that both choices work. Include contractor and agency hours when you direct and supervise the work, because that exposure belongs to your operation. Exclude them when a principal contractor or subcontractor reports their own figures, which avoids the same hours appearing in two rates. State the choice on the face of the report, and keep it unchanged between periods so your trend stays valid.
How do you calculate total hours worked for LTIFR?
Take actual hours worked by every person in your defined population over the reporting period, straight from payroll or the timekeeping system. Include overtime and double shifts at their real hour value. Exclude annual leave, sick leave, public holidays and other paid time when nobody worked, plus standby, on-call and home working with no exposure. Multiply headcount by average hours per week and by weeks as a cross-check, then reconcile to payroll.
What is a good LTIFR rate?
There is no universal pass mark, because rates depend heavily on trade, site conditions and how the hours are counted. What you can say is whether you are improving: track a rolling twelve-month rate and compare it with your own previous year and with peer companies using an identical definition, multiplier and scope. A falling rate paired with a rising near-miss report count usually signals healthier reporting, not worse safety.
How do I calculate the OSHA lost time incident rate?
Under 29 CFR 1904 the incident rate is recordable cases multiplied by 200,000 and divided by hours worked, all employees. DART does the same arithmetic but excludes recordable cases with no lost or restricted work day, so it is the closer US equivalent of a lost time rate. Keep your OSHA 300 log complete, enter hours for every employee on the same basis each quarter, and record restricted duty days separately. Expect about 200,000 hours per full-time employee per year as a sense check on your denominator.
Conclusion
How to calculate lost time injury frequency rate comes down to three numbers: lost time injuries, hours worked, and the multiplier you have chosen. Before the rate circulates, verify three things: the injury definition matches what your incident register actually recorded, the exposure hours cover the whole reporting period for the same population, and the multiplier is named on the page.
Once that holds, pair the rate with a severity measure and with leading indicators such as near-miss reports and critical control checks, since a frequency rate on its own tells you how often harm happened but nothing about how serious it was. For organizations tracking workforce participation alongside safety outcomes, our guide to calculating a wellness program participation rate shows the same normalisation logic applied to a different measure.