Skip to main content
Emergency lighting

BS 5266 Explained: Emergency Lighting Requirements

Emergency lighting is one of the few compliance regimes where the routine testing falls squarely on the building's own team rather than a contractor. This guide explains what BS 5266 requires, the difference between the monthly and annual tests, and the practical trap in the annual duration test that catches sites out.

Diagram summarising BS 5266 emergency lighting testing for facilities and estates teams, covering Where emergency lighting is required and Monthly and annual testing

In short

BS 5266-1 is the British Standard code of practice for emergency lighting of premises. It covers where emergency lighting is needed, how long it must operate for, and the testing regime: a short monthly function test and a full-duration test annually, both recorded in the premises log book.

Where emergency lighting is required

The purpose is to let people leave a building safely when the normal supply fails. That drives coverage: escape routes and the exits themselves, but also stairs, changes of level, direction changes, and the areas where a sudden loss of light would create its own hazard. Fire alarm call points and firefighting equipment need illumination so they can be found and used. Plant rooms, lift cars, and toilets above a certain size also commonly need cover.

Beyond escape lighting, the standard distinguishes standby lighting (letting normal activity continue) and high-risk task area lighting (letting a hazardous process be shut down safely). Most commercial buildings need escape lighting; the other two are driven by what happens in the building. As with fire detection, the coverage should follow from the fire risk assessment rather than from a generic layout.

  • Escape routes, exits, stairs, and changes of level or direction
  • At fire alarm call points and firefighting equipment
  • Plant rooms, lift cars, and larger toilet accommodation
  • High-risk task areas where a process must be made safe before evacuating

Monthly and annual testing

The monthly test is a short function test: simulate a supply failure long enough to confirm every luminaire and exit sign illuminates, then restore the supply and confirm the charging indicators are healthy. It is deliberately brief, because a long test flattens the batteries and leaves the building unprotected while they recharge.

The annual test is the full rated duration — normally three hours — and this is where sites come unstuck. The system must operate for the whole rated period and every luminaire must still be working at the end of it, which is the point of the exercise: a battery that lights a fitting for ten minutes passes the monthly test and fails the emergency. Because the batteries need up to 24 hours to recharge afterwards, the annual test should be scheduled when the building can tolerate a period of reduced protection, which usually means outside occupied hours. Both tests are recorded in the log book with the date, result, and any remedial action.

  • Monthly — short function test, confirm all luminaires and signs illuminate
  • Annually — full rated duration, normally 3 hours, all luminaires working at the end
  • Allow up to 24 hours recharge after the annual test
  • Record date, result, and remedial actions in the log book for both

Self-testing systems and record keeping

Automatic self-testing systems carry out the function and duration tests on a programmed schedule and log the results centrally, which removes the manual labour and the risk of a test being skipped in a busy month. They do not remove the duty: someone still has to review the results, act on failures, and be able to produce the records. A self-testing system whose fault reports nobody reads is worse than a manual regime, because it creates a false sense that the matter is handled.

Whichever route a site takes, the log book is the compliance artefact. It should record the system's design and rated duration, the test results, faults and their rectification, and any alteration to the installation. Alterations matter more than people expect: partitioning an open floor, changing a door swing, or relocating a call point can all put a luminaire in the wrong place, and the record of what was installed where is what makes that reviewable.

Compliance & CAFM

Track emergency lighting tests on a compliance register

PM Assist's building compliance software holds this duty on a register for every building you manage, included in every paid plan.

  • Emergency Lighting Test sits on the compliance register, showing when it was last done, when it is next due and the certificate behind it.
  • Duties with no record held are listed as gaps, so nothing is missed just because nobody logged it.
  • AI reads your certificates and proposes the dates and remedial actions; a person checks them against the source document before they count.

Sources

The regulatory positions on this page come from the following primary sources. Where guidance and legislation differ, the legislation governs — always confirm the requirements that apply to your own site and jurisdiction.

Frequently asked questions

How often should emergency lighting be tested?
Monthly and annually. The monthly test is a brief function test confirming every luminaire and exit sign illuminates on loss of supply. The annual test is a full-duration test — normally the full three hours the system is rated for — with all luminaires still working at the end. Both results go in the log book, along with any remedial work.
Why is the annual emergency lighting test three hours?
Because three hours is the rated duration of most systems, and the test exists to prove the batteries genuinely deliver it. A degraded battery will happily light a fitting for the few minutes a monthly test takes while failing badly short of three hours in a real evacuation. Only the full-duration test exposes that. Some systems are rated for one hour, in which case the annual test matches that rating.
Can I do the emergency lighting test myself?
Yes — the routine monthly and annual tests are commonly carried out by the building's own staff or FM team, and BS 5266 does not require a specialist for them. What matters is that whoever does it understands the system, records the results properly, and knows what a failure looks like. Remedial work on the installation itself is a different matter and needs a competent electrician.
Do self-testing emergency lights remove the need for manual testing?
They remove the manual testing labour but not the duty. An automatic system runs the function and duration tests on schedule and logs results, but someone still has to review those results, act on failures, and produce the records when asked. A self-testing installation whose fault reports go unread gives a false sense of compliance while leaving real failures in place.

Keep test records to hand

Emergency lighting log books, commissioning data, and layout drawings in one searchable library — so proving a year of tests takes seconds.

  • Upload and organise building documentation
  • AI-powered search across all your manuals
  • Source-cited answers for every query
  • Team collaboration and access control
  • No credit card required to start

Knowing the standard is one thing. Finding it in the paperwork is another.

Ask an O&M manual a question and get the answer back with the page it came from. No signup, and you can drop in your own PDF instead of the sample.

“When was the last F-Gas check?”

Runs against a sample chiller manual so you can see the shape of the answer — then try it on one of yours.