Workplace Deaths in Britain 2025/26: What HSE’s New Figures Actually Show

HSE published its provisional 2025/26 fatal injury figures on 1 July 2026. This analysis covers what they show, what the coverage will probably get wrong, and what the numbers mean for employers planning their next twelve months of risk control.

Every figure below is sourced to a named HSE data table or publication. Where we have done our own arithmetic on HSE’s published numbers, we say so.

The headline: 126 deaths, and why it is not comparable with last year

Source: HSE RIDFATAL Tables 1 and 4; RIDHIST Table 1.

The 126 figure will be widely reported as a fall from 128. Treat that comparison with caution, because it is not comparing like with like.

HSE publishes fatality counts as provisional first, then finalises them roughly a year later as late-reported deaths and reclassifications come in. Last year’s provisional figure was 124. The final figure was 128 — an increase of four worker deaths (and four members of the public) between the two counts. HSE’s own note on the current tables confirms 2025/26 will be finalised in July 2027.

So the honest reading is: 2025/26 provisionally stands at 126 against a finalised 128, and will very likely revise upward. Whether it ends up above or below 2024/25 is not yet knowable. Any coverage stating that workplace deaths “fell” this year is comparing a provisional number with a final one.

What can be said with confidence is the long-run picture. HSE records 126 deaths in 2025/26 against 217 twenty years ago in 2005/06 and 495 in 1981. Apart from pandemic-affected years, this is the lowest single-year count on record.

One further detail worth noting, because it is a change most readers will miss. HSE has revised its own trend language this year. Its previous assessment described the recent fatality rate as “broadly flat”. The 2026 report now describes a “downward trend, though the rate of decline has slowed in recent years”, explicitly calling this a refinement of previous assessments following updated modelling. The regulator’s position on whether Britain is still improving has quietly shifted.

Who died: the age finding

This is the most striking pattern in the 2025/26 data, and it is HSE’s own analysis rather than ours.

Workers aged 60 and over accounted for 32% of all worker fatalities in 2025/26, despite making up only 12% of the workforce.

Source: HSE RIDFATAL Table 4.

The rate data makes the point more sharply than the counts. On HSE’s five-year average, workers aged 60–64 have a fatal injury rate around twice the all-ages rate, and workers aged 65 and over a rate four times the all-ages rate. In 2025/26 the rate for the over-65s was 1.62 per 100,000 against an all-ages rate of 0.37.

That is a 2.7-fold over-representation of the over-60s relative to their share of the workforce — our arithmetic on HSE’s two published percentages.

There is a practical reading here for employers. An ageing workforce is not a neutral fact for risk assessment. If your workforce profile has shifted older over the past decade — and in construction, agriculture and haulage it very likely has — then a risk assessment written against a younger workforce profile is out of date. Manual handling capability, working at height, recovery from minor injury and reaction time in vehicle-movement areas all change with age, and the fatality data shows the consequence.

A caution on the source, for anyone checking our work: HSE’s summary web page currently presents this age table under a column header reading “2024/25” while the section heading says 2025/26. The underlying data table (RIDFATAL Table 4) is unambiguous — the 85/40/1 split belongs to 2025/26. We have cited the data table rather than the web page.

Who died: the self-employed finding

The second structural pattern is employment status.

Self-employed workers make up around 15% of the workforce but accounted for around a third of worker deaths in 2025/26 — 40 of the 126. On HSE’s five-year averages, the fatal injury rate for self-employed workers is 0.93 per 100,000 against 0.30 for employees, roughly three times higher.

The two findings compound. Self-employment is more common among older workers, and both agriculture and construction — the two deadliest sectors — carry high self-employment rates. The worker most at risk of dying on a British worksite is not the new starter; it is the experienced, older, self-employed contractor, often working alone or in a small gang, frequently outside any formal safety management system.

That is an uncomfortable finding for an industry whose safety culture is largely built around inducting employees.

How they died

Source: HSE RIDFATAL Table 3.

Falls from height remain the leading cause. HSE’s report notes that falls have been the main cause of work-related fatal injury in almost every year since at least 2001/02 — a quarter of a century in which the single largest killer has not changed.

On five-year averages, falls account for around 37 deaths a year (28% of the total), with struck-by-vehicle and struck-by-object each averaging 22 deaths a year (17% each). Those three causes alone account for roughly six in ten workplace deaths.

The sector concentration is instructive. Over the last five years, more than half of all construction deaths were falls from height, averaging 19 a year. Struck by moving vehicle accounted for 48% of deaths in waste and recycling, 30% in transportation and storage, and 23% in agriculture, forestry and fishing.

For most employers, this means the risk register writes itself. If you work at height, working at height controls are your single highest-value intervention. If you run a yard, a warehouse or a depot, workplace transport and banksman competence are. The data has said the same thing for twenty-five years.

Which sectors

Source: HSE RIDFATAL Table 1. Rates are per 100,000 workers.

Two ways of reading this table, and they give different answers.

By count, construction is the largest single contributor at 25 deaths — as it is almost every year, because it is a very large sector.

By rate, construction is not especially dangerous relative to the genuinely high-risk sectors. On HSE’s five-year averages, agriculture, forestry and fishing runs at 21 times the all-industry rate, waste and recycling at 10 times, and construction at around 5 times. Manufacturing, transport and storage, and admin and support services sit at roughly 1.5 to 2.5 times average.

Two year-on-year movements are worth flagging, with the provisional-versus-final caveat applied to both. Construction fell from 35 to 25, a 29% drop on our arithmetic — but 2024/25 was itself an unusually high year for construction, so this may be regression to the mean rather than improvement. Manufacturing moved the other way, from 11 to 18. Single-year moves in sectors with small absolute counts are noisy; neither should be over-read until the 2025/26 figures are finalised.

Separately, 104 members of the public were killed in work-related accidents in 2025/26 (provisional), up from a finalised 96 in 2024/25. Of these, 32 were on railways and 59 in service-sector industries.

The part nobody reports: illness dwarfs injury

Workplace deaths get the coverage. They are not where the harm is concentrated.

Source: HSE annual statistics 2024/25, published 20 November 2025. Our full breakdown of HSE’s annual statistics covers these figures in detail.

Set the 126 workplace deaths against 2,146 mesothelioma deaths from past asbestos exposure and the scale becomes clear. Occupational disease kills an order of magnitude more people than workplace accidents — it just does it decades later, which is why it does not generate headlines.

The working-days-lost split makes the same point:

Source: HSE working days lost statistics, 2024/25. Percentages are our arithmetic on HSE’s published figures.

Stress, depression and anxiety alone account for 55% of every working day lost to work in Britain. The average absence per case is 22.9 days, against 6.5 days for an injury.

Yet most organisations’ safety effort — inductions, toolbox talks, PPE, site inspections — is aimed almost entirely at the 11%. The mismatch between where safety attention goes and where working time is actually lost is the single largest inefficiency in UK health and safety practice, and it has been visible in the data for a decade. Mental health first aid provision and the HSE Management Standards approach to stress are no longer peripheral topics.

What it costs

Source: HSE Costs to Britain, 2023/24. Unit costs: £10,000 per injury case, £24,200 per ill-health case.

Two observations. First, the cost split mirrors the days-lost split — ill health is roughly three-quarters of the bill. Second, employers bear only £4.3bn of the £22.9bn, under a fifth. Most of the cost of unsafe work falls on the injured person and on the taxpayer, which is precisely why the activity is regulated rather than left to market incentives. Any board that evaluates safety investment purely against its own direct cost exposure is looking at 19% of the picture.

Which industries make people ill — a different list

The sectors that injure people and the sectors that make people ill are not the same sectors. This is one of the most under-appreciated points in the HSE data.

Statistically significantly above average for work-related ill health (2024/25): human health and social work (5,400 per 100,000), public administration and defence (5,530), education (4,940). All-industry average: 4,170.

Statistically significantly above average for non-fatal injury (2024/25): accommodation and food services (3,080 per 100,000), construction (2,500), transportation and storage (2,430), wholesale and retail including motor repair (2,250). All-industry average: 1,780.

Source: HSE industry statistics 2024/25. HSE notes no injury estimate was available for agriculture, forestry and fishing in this period; earlier periods show it significantly above average.

Note that only three sectors are statistically significantly above average on ill health — the confidence intervals on the others overlap the average, so describing them as “above average” would be unsound. We have listed only what HSE identifies as significant.

The pattern: public-facing, people-handling, emotionally demanding work makes people ill. Physical, fast-moving, vehicle-and-machinery work injures them. An employer in health, education or public administration who benchmarks their safety performance purely on injury rates is measuring the wrong thing for their sector.

A note on enforcement data

Anyone attempting to analyse HSE prosecutions or fines alongside these figures should know that HSE no longer publishes enforcement statistics as part of its annual statistical release. The regulator’s statistics site states that enforcement information is now treated as management information and appears only in HSE’s Annual Report and Accounts.

This is a meaningful gap. It means prosecution counts, conviction rates and fine totals no longer carry accredited official statistics status, and cannot be compared on a like-for-like basis with the injury and ill-health figures above. Any published analysis combining the two should say so.

What employers should take from this

1. Re-examine your risk assessments against your actual workforce age profile. The over-60s finding is the clearest signal in this year’s data. If your workforce has aged and your assessments have not been revisited, they no longer describe your risk.
2. Check how contractors and self-employed workers sit inside your safety management system. A three-times-higher fatality rate among the self-employed, concentrated in the sectors you probably work in, is not a statistic about somebody else.
3. Put working at height and workplace transport at the top of the register unless you have strong reason not to. Twenty-five years of data say these are what kill people. SMSTS and the NEBOSH Construction Certificate both address these directly for site management.
4. Rebalance effort toward ill health. Fifty-five per cent of lost working time is stress-related. If your safety programme does not have a proportionate mental-health component, it is aimed at the smaller problem.
5. Do not benchmark on the wrong metric for your sector. Use the ill-health rate if you are in health, education or public administration; use injury rate if you are in construction, transport, retail or hospitality.

Training is the mechanism most employers reach for, and it should follow the risk. IOSH Managing Safely equips line managers to run risk assessment properly at team level; the NEBOSH National General Certificate is the standard qualification for anyone owning the risk register. Where an organisation needs its own figures benchmarked against these sector rates, our consultancy team can scope that review. Call +44 (0) 3300 569534.

Methodology and how to cite this analysis

Data sources. All figures are from the Health and Safety Executive. Fatal injury data: RIDFATAL data tables and Work-related fatal injuries in Great Britain, 2026, published 1 July 2026. Ill health, injury, days lost and industry rate data: HSE annual statistics for 2024/25, published 20 November 2025. Cost data: HSE Costs to Britain, 2023/24. Historical series: RIDHIST data tables.

Status of figures. 2025/26 fatality figures are provisional and will be finalised in July 2027. 2024/25 fatality figures are final (revised). All ill-health, days-lost and industry-rate figures are for 2024/25 and are final. Cost figures are for 2023/24 and are a three-year rolling average.

Our own calculations. Where we have calculated percentages, year-on-year changes or ratios from HSE’s published numbers, this is stated in the text. We have not modelled, estimated or projected any figure. All rates per 100,000 workers are HSE’s own.

Geography. HSE fatality and injury statistics cover Great Britain, not the United Kingdom.

Citation. This analysis may be quoted and cited freely with attribution to KeyOstas and a link to this page. HSE data is reproduced under the Open Government Licence v3.0. For media enquiries or expert comment on these figures, contact KeyOstas on +44 (0) 3300 569534.

Last updated: 31 July 2026. We will update this analysis when HSE publishes the 2025/26 annual statistics in November 2026, and again when the fatality figures are finalised in July 2027.

Frequently asked questions

How many people died at work in the UK in 2025/26?

The Health and Safety Executive recorded 126 worker deaths in Great Britain in 2025/26. This figure is provisional and will be finalised in July 2027. A further 104 members of the public were killed in work-related accidents.

Is 126 workplace deaths a fall from last year?

Not reliably. The 126 figure is provisional, while the 2024/25 figure of 128 is final. Last year’s provisional count was 124 and was revised up to 128 when finalised, so 2025/26 is likely to rise as late reports come in. Provisional and final figures should not be compared directly.

What is the most dangerous industry in the UK?

By rate, agriculture, forestry and fishing, at 8.09 deaths per 100,000 workers in 2025/26 — around 21 times the all-industry average on HSE’s five-year figures. Waste and recycling is second at around 10 times average. Construction has the highest absolute number of deaths at 25 because it is a large sector, but its rate is around 5 times average.

What is the leading cause of workplace death in the UK?

Falls from height, which killed 31 workers in 2025/26. HSE notes falls have been the leading cause of work-related fatal injury in almost every year since at least 2001/02. Being struck by a moving vehicle at 24 deaths and struck by a moving object at 21 deaths are second and third.

Are older workers more likely to die at work?

Yes, substantially. Workers aged 60 and over accounted for 32% of worker deaths in 2025/26 but make up only 12% of the workforce. Workers aged 65 and over have a fatal injury rate around four times the all-ages average on HSE’s five-year figures.

Are self-employed workers more at risk than employees?

Yes. HSE’s five-year figures show a fatal injury rate of 0.93 per 100,000 for self-employed workers against 0.30 for employees — around three times higher. Self-employed workers are roughly 15% of the workforce but accounted for about a third of worker deaths in 2025/26.

How many working days are lost to work-related ill health and injury?

An estimated 40.1 million in 2024/25. Work-related ill health accounts for 35.7 million or 89% and workplace injury 4.4 million or 11%. Stress, depression and anxiety alone account for 22.1 million days, which is 55% of the total.

What does workplace injury and ill health cost the UK?

HSE estimates £22.9 billion for 2023/24. Ill health accounts for £16.4 billion or 72% and injury £6.5 billion or 28%. Individuals bear £13.4 billion, government £5.2 billion and employers £4.3 billion.

When does HSE publish workplace fatality statistics?

Provisional fatal injury figures are published in early July each year. The full annual statistics covering ill health, injuries, days lost and costs are published in November. Provisional fatality figures are finalised roughly two years later — the 2025/26 provisional figures will be finalised in July 2027.

Are workplace deaths in Britain falling?

Over the long run, yes. HSE recorded 126 deaths in 2025/26 against 217 in 2005/06 and 495 in 1981. HSE’s 2026 assessment describes a downward trend in the rate of fatal injury, while noting the rate of decline has slowed in recent years — a change from its previous description of the recent rate as broadly flat.