Between January 2015 and November 2025, employers in federal-OSHA-jurisdiction states filed 5,584 severe injury reports involving forklifts, hospitalizations, amputations, and eye losses that must be reported within 24 hours under 29 CFR 1904.39. Across the ten complete calendar years in that window, that is an average of roughly 515 severe forklift injuries reported every year, and the pace has barely moved in a decade: 465 reports in 2015, 562 at the 2018 peak, 461 in 2024. The dataset also carries 430 reports for 2025, but that year runs only to 30 November, so it is counted in the total above and left out of the annual average and every year-on-year comparison on this site.
How many forklift accidents happen per year?
Of those 5,584 reports, 983, about one in six, involved an amputation. Fingers and hands caught between the truck and racking, feet crushed under wheels, and limbs pinned in tip-overs recur throughout the incident narratives. On the fatality side, the Bureau of Labor Statistics counted 614 deaths in forklift-related incidents from 2011 through 2017, roughly 88 per year, and its injury survey attributes about 7,000–7,300 nonfatal forklift injuries with days away from work to a typical recent year.
Which states report the most forklift accidents?
Texas leads the covered states with 889 severe injury reports over the decade, 80 of them in 2024 alone, followed by Florida, Georgia, Ohio, and Pennsylvania. Rankings for every covered state, updated with each data refresh, are on the forklift injuries category page, and each state has its own page with the full yearly series: see Texas or browse the states index.
Two caveats make these numbers honest. First, this dataset covers establishments under federal OSHA jurisdiction; states running their own OSHA-approved plans, California, Washington, Michigan and others, report into their own systems and are not shown. Second, severe-injury reports are a reporting-based measure: research on the program suggests meaningful under-reporting, so true counts are higher, not lower. Both caveats are documented on the methodology page.
Year by year
The annual series is remarkably stable: 465 reports in 2015, then 491, 513, 562, 554, 491, 522, 556, 539, and 461 in 2024. No year in the decade dipped below 460 or rose above 570. The 2020 dip tracks the pandemic’s warehouse disruption; the rebound to 556 by 2022 tracks the e-commerce buildout that added thousands of new dock doors and, with them, new truck-pedestrian interfaces. A flat line across a decade of safety-technology progress is the strongest argument in this dataset: the controls that already exist, segregation, detection, engineered speed limits, are not being deployed at the aisles, docks, and blind corners where the injuries keep happening. That is a deployment gap rather than a technology gap, and deployment gaps are the kind a single budget cycle can close.
What do the trends say about prevention?
A decade of flat numbers is itself a finding: training-only safety programs have not bent the curve. The incident narratives point at the same handful of mechanisms, pedestrians struck in shared aisles, workers caught between truck and structure, tip-overs, and falls from raised forks. Those first two mechanisms are exactly what pedestrian detection systems and disciplined traffic segregation attack, which is why our Risk Scorer weighs pedestrian exposure and near-miss history hardest, the free score tells you which mechanism your site is most exposed to.
Using these figures
Safety budget proposals travel better with a denominator attached, because an abstract national number is easy to discount while a local one is not: “889 severe injuries in Texas since 2015” lands harder in a budget meeting than any national abstraction ever will. Every figure on this page traces to a committed snapshot of the federal dataset, refreshed on a documented cadence, so numbers you cite from here will not drift under you. For the regulatory context behind the reporting rule, see the OSHA forklift regulations guide; for per-state series, start at the states index. Estimates are labelled as such throughout.