Forklift Accident Statistics

"Nonroadway incident involving a motorized land vehicle" is BLS's label for a crash or an overturn that happens somewhere other than a public road: a warehouse aisle, a loading dock, a yard, a construction site. For forklifts that is mostly the truck going over onto its side or off an edge with the operator on it. It was the largest single group in 2017 at 20 deaths.

"Struck by powered vehicle, nontransport" covers the 13 people who were hit by the truck while it was not driving from place to place. Think of lift work done standing still: a load being placed, forks or a mast moving, someone standing where they could be pinned against a rack, a wall, or another truck. The forklift is what hit them, but it was working, not traveling.

"Struck by falling object or equipment" was 12 deaths, and it means something came off the forks or off a rack and landed on a person. That is a load that was not secured or was stacked past what it could hold, or a lift made over someone's head.

Falls to lower level, 11 deaths, is usually a person up in the air who should not have been: riding on the forks, standing on a pallet raised by the forks, or going off a dock edge. Pedestrian vehicular incidents, 9 deaths, is the one everyone pictures, a worker on foot struck by a forklift that was traveling (see our forklift pedestrian safety controls). The word "vehicular" is what separates it from the nontransport category above: here the truck was in transportation use when it hit someone.

OSHA has put a number on how much of this is a training problem, and it is not the number you have read. In the 1995 proposed rule for powered industrial truck operator training, 60 FR 13782, OSHA wrote that:

"powered industrial truck accidents cause approximately 85 fatalities and 34,900 serious injuries each year. It is estimated that approximately 20 to 25 percent are at least in part caused by inadequate training" ⁷

Read the qualifier: "at least in part." OSHA was not saying a quarter of forklift accidents have a single cause and that cause is training. It was saying training is somewhere in the chain for about one in four or five.

The "70 percent of forklift accidents are preventable with training" line that shows up on safety blogs and vendor sites comes from the same 1995 document, and it is a misreading of it. The 70 percent is a change in operator error rate from one training study OSHA cited: operators' mean error rate fell from .23 before training to .09 after, then to .07 during a later retention check, which is a 70 percent reduction in errors. OSHA closed off the leap to accident counts in the very next line of the same document:

"Although a 70 percent error rate reduction can not be directly equated to a corresponding reduction in the number of accidents that this or any other group of operators will experience..." ⁷

Anyone quoting 70 percent as an accident-prevention rate is quoting past OSHA's own caveat.

What OSHA settled on is smaller and more useful. In the 1998 final rule, 63 FR 66237, the agency wrote:

"Based on published reports, the Agency had estimated in the proposal that 44 to 77 percent of accidents could be avoided by compliance with the Agency's proposed rule. OSHA has since decided that a more conservative estimate of 25 percent of accidents more accurately reflects the percentage of accidents that will be averted by compliance with the final standard. This 25 percent reduction in fatalities applies to the Agency's estimated 42 fatalities each year that are potentially preventable, which results in an estimated 11 fatalities avoided each year under the final standard" ⁸

So OSHA's own, final, published position: training compliance prevents about 11 forklift deaths a year out of the 42 it considered preventable at the time. Twenty-five percent, not seventy.

Twenty-five percent is still the biggest single change an employer can make, and it is the one employers keep getting caught on. Powered Industrial Trucks, 29 CFR 1910.178, was the eighth most frequently cited OSHA standard in fiscal year 2025, the year running from October 1, 2024 to September 30, 2025 ⁹. Inspectors wrote 2,150 citations under it, a final count from OSHA's own information system reported by Safety+Health magazine ¹⁰; OSHA's public Top 10 page prints the ranking but not the number. The standard covers more than training, and neither OSHA nor Safety+Health breaks the count down by paragraph, so we cannot tell you how many of those 2,150 citations were for training specifically. What we can tell you is that paragraph (l) of that standard, operator training and evaluation, is a yes-or-no question about something the employer either did or did not do before handing someone the keys.