UK Workplace Fatality Statistics 2025/26 Explained: The Data Behind HSE's "Record Low"

Figure 1: Provisional worker fatalities by main industry sector for 2025/26, showing construction and agriculture continuing to lead absolute totals. Data source: HSE RIDDOR

HSE's UK workplace fatality statistics for 2025/26, published on 1 July, are being framed as a record low. The reality is more complicated. 126 workers died in Great Britain over the year. Excluding the years affected by the pandemic, HSE describes this as provisionally the lowest single-year total on record, down from 217 deaths twenty years ago and 495 in 1981. A companion international comparison places Britain among the safest countries in the world to work.

IOSH's response, from Ruth Wilkinson, struck a different note. She pointed out that fatalities have remained broadly unchanged for several years despite the long-term progress, and called it a systemic challenge requiring renewed focus.

"While Great Britain has made significant progress over the long term, it is deeply concerning that the number of work-related fatalities have remained broadly unchanged for several years. This suggests a deeper systemic challenge, one that requires a renewed focus and collective action to drive further improvements, with prevention-first approaches needed at the very heart of this.” Ruth Wilkinson, Head of Policy and Public Affairs, IOSH

Most coverage this month will report the record-low framing and stop there. The underlying data tells a more complicated story, and it's the story worth reading if you manage risk for a living rather than report on it.

Contents

• UK workplace fatality statistics 2025/26 at a glance

• 1. The "record low" claim doesn't survive contact with last year's figure

• 2. Count and rate tell two different stories about workplace deaths by industry

• 3. Self-employed worker deaths are the fault line getting the least attention

• 4. The age curve behind workplace fatalities is steeper than it first looks

• 5. Most public deaths don't happen where you'd expect

• 6. The long-term decline in HSE fatality statistics has stalled

• Where the Next Reduction Has to Come From

UK workplace fatality statistics 2025/26 at a glance

→ 126 workers killed in Great Britain in 2025/26, per HSE

→ 104 members of the public killed in work-related incidents, up from 92 the year before

→ Self-employed workers account for roughly 40% of worker deaths despite making up around 15% of the workforce (2024/25 figures, the most recent published at this granularity)

→ Workers aged 60 and over account for around a third of deaths despite making up roughly 12% of the workforce

→ Falls from height remain the single largest cause of death, at 31 fatalities

1. The "record low" claim doesn't survive contact with last year's figure

HSE's framing compares 2025/26 against 2005/06 and 1981. It doesn't mention that 2024/25's figure was 124, lower than this year's 126. Provisional figures do get revised once finalised, so 2024/25 may yet move, but as published, the record-low claim sits alongside an actual year-on-year rise of two deaths. HSE isn't hiding this; both figures are on its own site, but the press release frames the comparison against twenty and forty-five years ago rather than against last year, and that's the comparison most coverage will repeat without checking.

2. Count and rate tell two different stories about workplace deaths by industry

Construction (25 deaths) and agriculture, forestry and fishing (22 deaths) top the list of worker deaths by industry, and that's the version most coverage will run with. Taken alone, it's a misleading guide to where the individual risk actually sits.

By fatality rate, HSE's own figures (based on the five-year average to 2024/25) show agriculture, forestry and fishing running at around 22 times the all-industry rate, the highest of any sector. Waste and recycling comes next, at around 8 times the average. Construction, by contrast, sits at around 5 times the average, well below both, despite generating the largest headline count.

A worker in waste and recycling faces meaningfully higher personal risk than a construction worker, even though construction tops the table by count. Absolute counts track the size of an industry as much as its danger. Rate is the figure that describes the job itself.

Figure 2: Falls from height continue to rank as the single largest cause of fatal injuries in British workplaces, highlighting why absolute data numbers require deeper regulatory scrutiny

3. Self-employed worker deaths are the fault line getting the least attention

This is the statistic that deserves more scrutiny than it gets. Self-employed workers make up around 15% of the UK workforce and account for roughly 40% of worker deaths, a pattern that has held consistently across recent years (HSE hasn't yet published this specific breakdown for 2025/26, so this is based on the most recent detailed figures, for 2024/25).

The clearest verified example sits within agriculture. There, the fatal injury rate for self-employed workers is 11.72 per 100,000, against 5.86 for employees, exactly double, comparing like with like within the same industry. In construction, self-employed workers accounted for close to 45% of fatalities over the five years to 2024/25, despite being a smaller share of the construction workforce than that.

Almost every piece of health and safety infrastructure in the UK assumes an employment relationship underneath it: risk assessments, method statements, site inductions, supervision structures. Remove that relationship, as happens the moment someone is engaged as a sole trader or subcontractor, and the infrastructure has nothing left to attach to. Nobody is formally responsible for checking whether that person has had adequate rest, feels able to raise a concern, or is under commercial pressure to cut a corner they'd otherwise flag.

Construction routes labour through several tiers of subcontracting before it reaches the person doing the work. Each tier adds distance between whoever sets the safety standard and whoever is exposed to the risk. This gap sits in the structure of how work is contracted, and it's largely invisible to a regulatory model built around direct employment.

4. The age curve behind workplace fatalities is steeper than it first looks

Workers aged 60 and over account for around a third of all fatalities, against roughly 12% of the workforce. Reported alone, that reads as a demographic footnote. The shape underneath is sharper.

The fatality rate doesn't rise in a straight line with age. Based on HSE's most recent age-banded analysis, the rate for 60-to 64-year-olds runs at roughly double the all-ages average, and for workers 65 and over, it runs at roughly four times the average.

With the state pension age continuing to rise and more people staying in physically demanding trades for longer, this pressure is likely to build over the next decade rather than hold steady. Managing age as one line in a generic risk assessment won't keep pace with a workforce that's ageing this quickly in manual roles.

5. Most public deaths don't happen where you'd expect

104 members of the public died in work-related incidents in 2025/26, up from 92 the year before. Of those, HSE's own breakdown attributes only 13 to the traditional high-risk industries people picture when they think of workplace danger: seven in agriculture, forestry and fishing, three in construction, and three in waste and recycling. That leaves roughly 91 public deaths, the large majority, in what HSE classifies as the service sector: rail, retail, health and social care premises, leisure and hospitality, among others.

This isn't a one-off. In 2023/24, the equivalent split was 78 public deaths in the service sector against 9 in other industries, virtually the same pattern. The public mental image of "workplace danger" runs to hard hats and heavy plant. The data says the greater share of harm to non-workers is happening somewhere else entirely, in settings that don't get anywhere near the same safety scrutiny.

6. The long-term decline in HSE fatality statistics has stalled

Figure 3: While acute workplace accidents receive mainstream media coverage, occupational disease data highlight that historical exposures remain the largest driver of work-related mortality in the UK.

HSE's own long-term data describes the rate of decline as having broadly flattened over roughly the last ten years, once pandemic years are set aside. Set against 1981's 495 deaths, the figure of 126 invites a story of continuous improvement. The more accurate read is that the major gains were made in previous decades, largely through engineering controls and regulatory change, and that progress has essentially stalled since the mid-2010s.

This is the point IOSH's Ruth Wilkinson was making, and it lines up with HSE's own trend commentary. A system that made rapid early gains and has since stalled is usually telling you the interventions that worked against the easiest risks have run their course. What's left requires a different kind of intervention.

Where the Next Reduction Has to Come From

The long-term reduction from 495 deaths in 1981 to 126 last year reflects decades of genuine, hard-won progress, and that progress is real. It's also run out of road. A decade-long plateau, a self-employment gap that concentrates catastrophic risk in the workers least protected by the system, and a public death toll concentrated somewhere the public conversation never looks, all point in the same direction: the next reduction has to come from somewhere the current model doesn't reach.

Self-employment is the clearest starting point. Psychological safety, like any other layer of protection, can't be bolted onto a relationship that doesn't structurally exist. Employers, principal contractors and consultants working across subcontracted supply chains need to treat the absence of an employment relationship as a risk factor in its own right, and build oversight, communication and welfare support into contracts and site management specifically to close that gap. Systems designed for direct employees don't cover it by default.

This is where most H&S consultancy stops. A risk assessment gets updated, a policy gets amended, and the underlying relationship, the one where nobody's formally responsible for how a self-employed worker is actually doing, stays exactly as it was. Closing that gap means building oversight and welfare support into how contracts and site management work day to day, not adding another document to the pile. That's a longer job than a compliance review, and it's the one that closes the gap this piece keeps coming back to.

Sources: HSE, Health and Safety Executive Media Centre, "Latest annual work-related fatalities published," 1 July 2026. HSE, Work-related fatal injuries in Great Britain, statistics overview, 2026. HSE, Work-related fatal injuries in Great Britain, 2024/25 report. IOSH, media statement from Ruth Wilkinson, Head of Policy and Public Affairs, July 2026.

About the author

Beth Slade is the founder of Beth Slade Safety Ltd, a health and safety consultancy built around embedding psychological safety into core H&S practice rather than treating it as an add-on. She holds a NEBOSH Diploma, is a Certified member of IOSH, and is an ISO 45003 auditor, with over 20 years' experience across the industry. Connect with her on LinkedIn.