Psychosocial risk is the risk to health arising from how work is organised, from social factors at work, and from the work environment and the tasks themselves; in Great Britain it is not a separate legal category but part of the ordinary duty to assess and control risks to employees' health under regulation 3 of the Management of Health and Safety at Work Regulations 1999.
The term is newer than the duty. That is most of what makes it confusing, and it is why this page starts with what the words mean.
ISO 45003, the international guidance standard, defines psychosocial risk as the "combination of the likelihood of occurrence of exposure to work-related hazard(s) of a psychosocial nature and the severity of injury and ill-health that can be caused by these hazards", and groups those hazards into three categories: aspects of work organisation, social factors at work, and the work environment, equipment and hazardous tasks.
HSE does not use the phrase much. Its own framework, the Management Standards, covers the same ground under six headings: demands, control, support, relationships, role and change. The two vocabularies describe the same territory, and for practical purposes in Great Britain the six areas are the more useful set, because they are free, quotable and written for this regulatory context. They are explained in the six HSE Management Standards.
One definition worth borrowing from the regulator, because it keeps the whole subject on the right side of a line that matters. HSE defines stress as "the adverse reaction people have to excessive pressures or other types of demand placed on them", and states: "Stress is not an illness, but it can make you ill."
Psychosocial risk is a property of work. It is not a description of anybody's health.
Under the parent duty. This page is a wing of the employer duty of care, not a separate obligation.
Section 2(1) of the Health and Safety at Work etc. Act 1974 requires employers to ensure, so far as is reasonably practicable, the "health, safety and welfare at work" of employees. Regulation 3(1) of the 1999 Regulations requires a suitable and sufficient assessment of the risks to their health and safety, with no headcount threshold. Regulation 3(6) requires employers with five or more employees to record the significant findings, and any group identified as being especially at risk, in writing.
Neither instrument contains the words "psychosocial", "stress" or "mental health". The duty reaches this territory because both are framed around health without qualification, and because HSE states the position directly: "Employers have a legal duty to protect workers from stress at work by doing a risk assessment and acting on it. This is the same duty you have to protect people from other health and safety risks."
The civil route runs alongside it and is where most employers' real exposure sits. Under Hatton v Sutherland [2002] EWCA Civ 76, liability for psychiatric injury turns on whether harm to that particular employee was reasonably foreseeable, judged on what the employer knew or ought reasonably to have known (proposition (2), as reproduced in Easton v B&Q plc [2015] EWHC 880 (QB) at [50]).
These are Great Britain duties. The 1974 Act does not extend to Northern Ireland, and the 1999 Regulations extend to England, Wales and Scotland only.
This is the section most pages on this subject skip, and skipping it is why so many of the numbers in circulation are wrong.
The headline figures are solid but they are self-reports. HSE estimates 964,000 workers in Great Britain with work-related stress, depression or anxiety in 2024/25, a rate of 2,770 per 100,000, and 22.1 million working days lost to it, 22.9 days per case (published 20 November 2025). Those come from the Labour Force Survey and count conditions workers themselves attribute to their work. They are not diagnoses and not a clinical prevalence rate.
The trend is not the story it is usually told as. The rate stepped up once, at the pandemic, from 1,810 per 100,000 in 2018/19 to 2,480 in 2019/20. It then moved with no statistically significant year-on-year change for four years, and 2023/24 at 2,290 was lower than 2021/22 at 2,800, before rising significantly to 2,770 in 2024/25. HSE's own published wording for the longer run is that rates "had shown signs of increasing". It has never called this a crisis, and neither should anything you present.
Absences are not getting longer. Days lost per case has fallen from around 28.5 in the early 2000s to 21.3 averaged across 2022/23 to 2024/25. Total days lost are at a series high because there are more cases, not because each lasts longer.
The survey underneath all of it is under strain, and HSE says so. Labour Force Survey response rates fell from around 50% in early 2010 to below 40% at the pandemic onset, then "dropping to less than 25%", with recent quarters "rising above 25%". About a third of responses are given by proxy, and proxies under-report: spouse or partner proxies are 26% less likely to report a work-related illness, with no adjustment made. That last point cuts toward under-statement, which is worth saying out loud if anyone tries to use the caveat the other way.
The whole 2018/19 to 2023/24 series was revised in November 2025 when ONS reweighted the survey. Any figure from those years published before that date is superseded.
Nobody has current data on causes. HSE's published breakdown puts workload predominant, then lack of managerial support, then violence and bullying, then organisational change and role uncertainty. The survey period is 2009/10 to 2011/12, and HSE has retired the table from its live site to the National Archives. There is no current-year causes breakdown at all. Anyone quoting a percentage for what causes work-related stress is quoting data that is thirteen to sixteen years old, whether they say so or not.
And there is no cost figure for this. HSE publishes a total cost of workplace injury and new cases of work-related ill health in Great Britain of £22.9 billion for 2023/24, but the word "stress" appears nowhere in that 42-page report, and its breakdown is by days off work rather than by condition. The large numbers that circulate for the cost of poor mental health to employers come from consultancies that sell workplace wellbeing services, are models, not measurements, and in at least one case contradict themselves on their own baseline within a single page. We do not use them.
One correction that runs directly against how this market sells. HSE does publish a workplace-size breakdown for stress, and it says the opposite of the intuitive claim: workplaces with fewer than 50 people have a statistically significantly lower rate of work-related stress, depression or anxiety than the all-sizes average, and workplaces of 250 or more a significantly higher one, across 2022/23 to 2024/25. That is workplace size, not company size; "small" includes the solo self-employed, and it is not evidence that anyone manages the risk better; sector and occupation mix confound it and HSE offers no causal explanation.
Nothing on this list is an obligation, and each is sold as one somewhere.
There is no duty to provide counselling or an employee assistance programme, and providing one is not a defence: the Court of Appeal held in Intel Corporation (UK) Ltd v Daw [2007] EWCA Civ 70 at [45] that "the reference to counselling services in Hatton does not make such services a panacea by which employers can discharge their duty of care in all cases". There is no duty to run a survey, including HSE's own. There is no duty to follow the Management Standards, and no score to hit against them. There is no duty to adopt ISO 45003, and no accredited certification to it exists. There is no duty to appoint mental health first aiders. Work-related stress is not reportable under RIDDOR. And no format, template or retention period is prescribed for the assessment anywhere.
What is required is the assessment, the written record at five or more employees, action on what it finds, and review.
Two things are moving and neither of them is a new duty.
HSE has said, in its Business Plan 2026 to 2027 published 28 May 2026, that it is "developing our regulatory approach to reduce work-related stress by prioritising preventive measures", with a Q4 deliverable to "deliver the evidence required to develop regulatory interventions". Work-related stress is a named priority in that plan and in the 2022 to 2032 strategy. No new regulation, code of practice or consultation has been announced, and the plan's stated inspection investment goes to asbestos, noise, musculoskeletal and respiratory risks.
The Keep Britain Working review, published 5 November 2025, recommends a voluntary certified standard and better workplace health provision rather than new statutory duties. Note that the review does not use the word "psychosocial" or the word "stress" anywhere in it, so it is not evidence about where stress regulation is going, whatever it gets cited for.
As at August 2026 there is no announced legislative proposal in Great Britain to create a specific psychosocial risk duty. The direction of travel is evidence-building by the regulator and voluntary standard-setting elsewhere.
If you need to know what a defensible assessment actually looks like, and what an inspector or a solicitor asks for, that is psychosocial risk assessment: what HSE actually expects.
If you want the six areas explained with HSE's own wording, the six HSE Management Standards.
If someone has asked about the international standard, ISO 45003, plainly explained.
If you are choosing between an annual instrument and something continuous, annual survey or continuous evidence.
And if you want the whole legal position, the employer duty of care.
The free strategy audit takes twenty minutes and gives you your gaps in writing. No card, no call booked at the end of it unless you ask for one.
Start the free audit →