How often should legionella testing be done?
On this page
Key points
- Monitoring has fixed frequencies: weekly flushing, monthly sentinel temperatures, quarterly cleaning tasks, annual inspections — set out in HSG274 Part 2 Table 2.1.
- Laboratory sampling has no fixed interval for most hot and cold water systems: it is triggered by the risk assessment, not the calendar.
- Exceptions with routine sampling built in: cooling towers (quarterly baseline, HSG274 Part 1), spa pools (at least monthly), healthcare (schedule set by the water safety plan under HTM 04-01).
- The five triggers that bring a test forward: commissioning, doubt about control, post-positive verification, after disinfection or shutdown, and a linked case.
- A lab result is a snapshot of one day. The monitoring log is the evidence of control over time — keep records for at least five years.
How often legionella testing should be done has a two-part answer, because "testing" covers two different activities. Routine monitoring runs on fixed frequencies: weekly flushing of little-used outlets, monthly temperature checks at sentinel outlets, quarterly cleaning tasks and annual inspections, as set out in HSG274 Part 2 Table 2.1. Laboratory sampling has no fixed interval for most hot and cold water systems at all: it is carried out when the risk assessment identifies a need. Only higher-risk systems — cooling towers, spa pools, healthcare premises — have routine sampling built into their control regime.
The monitoring frequencies: the HSG274 Table 2.1 matrix
HSG274 Part 2 is explicit (paragraph 2.79) that the frequency of inspection and monitoring depends on the system's complexity and the susceptibility of its users, and that the risk assessment should define it — but its Table 2.1 gives the standard starting checklist for hot and cold water systems. Consolidated into one matrix, the routine frequencies look like this:
| Frequency | Task | Target |
|---|---|---|
| Weekly | Flush infrequently used outlets (anything unused for seven days or more) and log each flush | Run until the temperature at the outlet stabilises; see how often outlets should be flushed |
| Monthly | Hot water temperatures at sentinel outlets (nearest, furthest, long branches) | At least 50°C within one minute (55°C in healthcare) |
| Monthly | Calorifier flow and return temperatures | Flow as close to 60°C as practicable without going below; return not below 50°C |
| Monthly | Cold water temperatures at sentinel taps | Below 20°C within two minutes of running |
| Quarterly | Subordinate-loop return temperatures on circulating systems (ideally a rolling monthly rota) | Above 50°C within one minute (55°C in healthcare) |
| Quarterly | Dismantle, clean and descale shower heads, inserts and hoses | Or more often as fouling or risk factors indicate |
| Annual | Inspect cold water storage tanks; check stored and incoming temperatures (summer) | Remedial work where necessary; temperatures per the profiling |
| Annual | Inspect calorifiers internally and clean by draining; purge base debris where there is no hatch | Or as the rate of fouling indicates |
| Annual | Check thermal insulation is intact | Weatherproof exposed components |
| Rotational | Temperatures at a representative selection of other outlets | Build a temperature profile of the whole system over a defined period |
These are the starting frequencies, not the ceiling: the risk assessment tightens them for higher-risk systems and more susceptible users, and the chosen schedule belongs in the written scheme of control. The monitoring schedule page builds the same stack by system type, and the temperature checker structures the monthly round.
The sampling frequencies: risk-based, with three standing exceptions
For laboratory sampling, HSG274 Part 2 sets no routine interval for hot and cold water systems, because temperature — not sampling — is the primary control. Sampling frequency by system type:
| System | Routine sampling frequency |
|---|---|
| Domestic-style hot and cold water (rented homes, small offices) | None. Temperature monitoring and flushing; sample only on a trigger. |
| Larger or complex hot and cold systems (hotels, schools, workplaces) | No fixed interval; the risk assessment sets any schedule, and triggers bring a test forward. |
| Healthcare and augmented care | Set by the water safety plan under HTM 04-01, reflecting susceptible occupants. |
| Evaporative cooling towers and condensers | Quarterly baseline under HSG274 Part 1, alongside weekly dip slides; tightened to weekly while control is being re-established. |
| Spa pools and hot tubs | At least monthly under HSG274 Part 3 and HSG282, alongside continuous disinfectant monitoring. |
The consistent pattern: the more a system relies on chemistry rather than temperature, and the more susceptible the people exposed, the more sampling moves from "on a trigger" to "on a schedule".
The triggers that bring a test forward
Where there is no standing schedule, five events should put sampling on the agenda. Commissioning a new or substantially altered system, to confirm it starts clean. Doubt that control is working — persistent out-of-range readings on the monthly round with no fault found. Verification after a positive result, to confirm remediation worked. After disinfection or a long shutdown, before normal use resumes. And a suspected or confirmed case of Legionnaires' disease linked to the premises. How often water should be tested works through these triggers in detail, and sampling and action levels explains what the 100 and 1,000 CFU per litre results mean when they arrive.
Monitor on a calendar; sample on a reason. The weekly flush and the monthly temperature round are the control — a laboratory test answers the specific question those checks raise.
Landlords and small premises
For a typical rented home or small office, the practical answer is that routine laboratory testing is not expected at any frequency. HSE landlord guidance treats simple domestic systems as controllable by temperature and basic management: hot water stored hot, cold kept cold, outlets in regular use, the system flushed through after a void, heads and hoses clean. What the landlord must be able to show is the written risk assessment and the records of those controls — not a laboratory certificate. Anyone selling a blanket "annual legionella test" for a simple domestic system is selling a snapshot that, on its own, neither controls the risk nor discharges the duty.
Why the records outrank the results
A negative sample describes one outlet on one day; it says nothing about next month. The monitoring log, kept for at least five years, is what demonstrates continuing control to an inspector, an insurer or a court — and the monitoring frequencies above only count if they are carried out consistently and written down. That is the honest shape of "how often": the calendar tasks every week, month, quarter and year, sampling whenever a trigger or the assessment calls for it, and all of it evidenced in the log book.
Free legionella risk assessment template
A structured Word document following the five-step approach in ACOP L8. Covers risk identification, written scheme, monitoring, and records. If it isn't written down, you can't evidence it.
Follows ACoP L8 and HSG274 Part 2. Free. No spam.
Free legionella risk assessment template
A structured Word document following the five-step approach in ACOP L8. Covers risk identification, written scheme, monitoring, and records. If it isn't written down, you can't evidence it.
Follows ACoP L8 and HSG274 Part 2. Free. No spam.
Frequently asked questions
Is there a legal requirement to test for legionella every year?
No. There is no statutory or ACOP interval requiring routine laboratory legionella testing of ordinary hot and cold water systems — annual or otherwise. The routine duties are monitoring tasks with set frequencies under HSG274 Part 2 Table 2.1: weekly flushing of little-used outlets, monthly sentinel temperature checks, periodic inspections. Laboratory sampling is added when the risk assessment identifies a need. Cooling towers are the standing exception, with quarterly sampling as the baseline under HSG274 Part 1.
How often should a landlord test for legionella?
Routine laboratory testing is not normally expected in rented homes. HSE landlord guidance treats simple domestic systems as controllable by temperature and basic management: hot water hot, cold water cold, outlets in regular use, the system flushed through after a void. The evidence a landlord needs is the written risk assessment and the records of those controls, not a laboratory certificate. Sampling enters the picture only where the assessment finds a specific reason, such as a system that will not hold temperature.
How often should cooling towers be tested for legionella?
Quarterly is the general baseline for routine legionella sampling of cooling towers under HSG274 Part 1, sitting alongside weekly dip-slide trend checks of the treatment programme. The frequency is increased — typically to weekly — during commissioning, after a positive result, and while a system that has lost control is being brought back into line. The exact schedule for a given tower is set by its risk assessment and written scheme of control.
What is the difference between legionella monitoring and testing?
Monitoring is the routine checking you do yourself — temperatures, flushing, visual inspections — at fixed weekly, monthly, quarterly and annual frequencies. Testing is laboratory analysis of a water sample, which tells you whether legionella was actually present in that water on that day. Monitoring runs on a calendar; testing runs on reasons. The monitoring log, not any single lab result, is what evidences continuing control.
Related water hygiene products and services from trusted UK providers will appear here.