A Legionella risk assessment is not a document you finish. It is a claim about your water system that stays true only until something changes. The awkward part of “when to review” is that the trigger that matters most is rarely a date in the diary. It is an event — a refit, a tenant moving in, a wing left empty, a sample that drifts — that quietly makes yesterday’s assessment wrong.
UK guidance is built around exactly that. It expects you to review regularly, but it pins the real obligation on a simpler question: is the assessment still valid? Lose sight of that and you end up running a tidy review cycle that sails straight past the calorifier you swapped last spring.
So treat review as two jobs running side by side. There is the scheduled look that catches slow drift, and the event-driven check that catches change. The second one is what protects people, and it is the one most sites forget.
What “review” actually means in UK guidance
Under HSE’s L8 Approved Code of Practice, the risk assessment sits at the head of a management chain — assess the risk, write a scheme of control, appoint a competent responsible person, implement it, monitor it and keep records — and every link depends on the assessment still describing reality [1]. HSE summarises the core duties as identifying and assessing sources of risk, managing and controlling them, keeping records and reviewing what you have done [2]. If the assessment underneath that scheme has gone stale, the records on top of it are documenting control of a system that no longer exists.
L8 frames review around validity rather than a fixed clock. The expectation is that you review the assessment regularly and, specifically, whenever there is reason to believe it is no longer valid [1]. BS 8580-1, the British Standard code of practice for Legionella risk assessment, is the document many assessors work to when they carry that review out [3].
That word “valid” carries the weight. An assessment is valid while the system it describes, the people exposed to it, and the controls keeping it safe are all still as written. Change any of those and the question stops being “is it overdue?” and becomes “is it still true?”
Many duty holders also set a default review interval — often cited as around two years — so nothing slides indefinitely. Treat that as a backstop, not the rule. Your own assessment should state its own review frequency, and the genuine trigger in guidance is loss of validity, which can land the week after your last review [1].
The trigger test: does this change mean a review?
When something changes on site, run it through a short test before you decide whether to act. Work top to bottom. The first “yes” means review.
- Has the water system itself changed? New calorifier, added or rerouted pipework, a removed or capped outlet, altered storage, a different treatment regime → review the affected parts now.
- Has the use of the building or its water changed? A floor brought back into service, a department mothballed, occupancy that has dropped or spiked, a change of tenant → review, because exposure and stagnation patterns have moved.
- Has the population exposed changed? New vulnerable occupants — a care use, a nursery, a clinical area introduced into a previously ordinary building → review with their susceptibility in mind.
- Is your monitoring telling you the controls are not holding? Temperatures repeatedly out of range, recurring positive samples, flushing routinely missed → review the assessment, not just the failing task.
- Has new information emerged? Revised guidance, a manufacturer safety notice, a finding from a similar site you operate → review against it.
- Has a case of Legionnaires’ disease been confirmed or suspected in someone exposed to the system? → review immediately, and check whether reporting duties under RIDDOR apply [4].
- None of the above, but you have reached your stated review date? → carry out the scheduled periodic review anyway.
If you can answer everything except the last line with “no” and the diary date has not arrived, you are entitled to leave the assessment alone — but write down that you checked. A reviewed-and-unchanged decision is itself a piece of evidence.
What these triggers look like in a real building
The triggers read clean on paper and messy on site. A few that catch people out:
A refurbishment that “didn’t touch the water.” A contractor reroutes a basin feed and leaves a metre of capped dead pipe behind a new stud wall. The system changed; the assessment did not. That dead leg is now an unassessed pocket of stagnation.
The wing that came back. A school shuts a block over a long summer, or an office floor sits idle after a downsizing and is later relet. Reoccupation is a textbook review trigger — months of still water are about to feed live outlets that people will actually use.
The quiet personnel change. The named responsible person leaves and nobody formally re-appoints. Competence and ownership are part of what the assessment assumes; losing them undermines it even though no pipe moved.
Drift in the monitoring data. Three months of a hot return arriving below target is not just a maintenance ticket. It is a signal that the control assumptions in the assessment may no longer hold, and recurring positive results in particular need reading in context rather than another round of re-sampling. The pattern across all of these is the same: a review trigger is any change to the system, its use, the people, or the evidence that the controls work.
Review and full reassessment are not the same job
A review is not automatically a fresh assessment. Most triggers call for a focused look at the part that changed — the new pipework, the relet floor — and a dated note confirming the rest still stands. A full reassessment is warranted when the change is broad or the existing document has decayed past patching: major refurbishment, a change of building use, or an assessment so out of date that nobody trusts what it says any more.
Getting that proportion right is where competence shows. Reassessing the whole estate after a single outlet change trains people to treat review as box-ticking; patching a major refit with a one-line note leaves a gap an inspector will find. The responsible person’s task is to size the response to the trigger, then record why they sized it that way — which is exactly the kind of decision your written scheme of control should make easy to trace.
Where this stops being a checklist
The judgement about when an assessment has stopped being valid — and how deep a review a given change deserves — is a competent-person call made on your specific system. A generic timetable cannot settle it for you. The triggers above, and the two-year figure, describe common UK practice; the binding answer for your site comes from your own risk assessment and the people accountable for it. And if a change has left you genuinely unsure whether a review is due, that uncertainty is usually the answer.
FAQ
Is there a legal duty to review the assessment every two years?
The two-year interval is a widely used default, not a fixed statutory deadline. L8 expects review on a frequency you set and, more importantly, whenever there is reason to believe the assessment is no longer valid [1]. A change to the system can make a review due long before any calendar interval expires.
If monitoring finds a problem, do I need a brand-new risk assessment?
Usually not a new one — but you do need to review. Repeated out-of-range temperatures or positive samples are evidence that the control assumptions may have failed, so revisit the relevant part of the assessment, decide whether the controls themselves need changing, and record the outcome rather than just closing the maintenance job.
Does a building standing empty trigger a review?
The return to use is the trigger to watch. A long void period changes stagnation and exposure patterns, so before reoccupation review how those outlets will be brought back safely — extended low or no use is one of the clearest reasons to look again at an assessment written for normal occupancy.
Your next step
Open your current risk assessment and find two dates: the review date it states, and the date it was actually written. Then list every material change to the building, its water system or its occupancy since that writing date. If the list is not empty, you already have your review scope — start with whichever item raises exposure most. And if the document names no review trigger at all, closing that gap is the first review you owe it.
Sources
[1] HSE, “Legionnaires’ disease. The control of legionella bacteria in water systems - Approved Code of Practice and guidance (L8)”. https://www.hse.gov.uk/pubns/books/l8.htm [2] HSE, “Legionnaires’ disease - what you must do”. https://www.hse.gov.uk/legionnaires/what-you-must-do/index.htm [3] BSI, “BS 8580-1:2019 - Risk assessments for Legionella control. Code of practice”. https://knowledge.bsigroup.com/products/water-quality-risk-assessments-for-legionella-control-code-of-practice-1 [4] HSE, “RIDDOR - Reporting of Injuries, Diseases and Dangerous Occurrences Regulations”. https://www.hse.gov.uk/riddor/