HSG38 lighting levels, and how to measure against them.

HSG38, Lighting at work, is the Health and Safety Executive's guidance on workplace lighting. For five kinds of work it recommends two figures: an average for the area as a whole, and a minimum for any single point in it. This page sets out HSE's table as HSE prints it, explains why both figures matter, and describes how to measure an area against them.

The table

HSE's recommended illuminances for different types of work, from the table that follows paragraph 94 of HSG38 (page 28 of HSE's free PDF). The rows are set out as HSE prints them.

HSG38 recommended illuminances, by activity. Figures in lux.
Activity Typical locations/types of work Average illuminance (lux) Minimum measured illuminance (lux)
Movement of people, machines and vehicles(a) Lorry park, corridors, circulation routes 20 5
Movement of people, machines and vehicles in hazardous areas; rough work not requiring any perception of detail Construction site clearance, excavation and soil work, loading bays, bottling and canning plant 50 20
Work requiring limited perception of detail(b) Kitchens, factories assembling large components, potteries 100 50
Work requiring perception of detail(c) Offices, sheet metal work, bookbinding 200 100
Work requiring perception of fine detail(d) Drawing offices, factories assembling electronic components, textile production 500 200

HSE's notes to the table, in our words:

Why there are two figures, and why the minimum matters

Each row gives an average for the work area as a whole and a minimum for any position within it. HSG38 gives the reason: an area can reach its average while parts of it stay darker, and those darker parts may put people at risk. The minimum is the lowest level HSG38 recommends anywhere in the area for health and safety.

So an area is measured against both. Take office desk work, which is work requiring perception of detail. The readings might average well over 200 lux while one corner reads 80 lux. That area is at or above HSG38's average figure and below its minimum figure, and a survey should say both.

HSG38's rows describe activities, so choose the row by the work done in the space rather than by what the room is called.

What the law says, and what HSG38 is

The legal duty is in regulation 8 of the Workplace (Health, Safety and Welfare) Regulations 1992, which apply in England, Wales and Scotland: every workplace must have suitable and sufficient lighting. The regulation gives no figures.

HSG38 is HSE's guidance on lighting at work. It is guidance, not law. The edition HSE publishes is the second, from 1997. Its figures are HSE's view of the minimum light needed for health and safety. They are not the levels a lighting designer works to today, and they are not a design target.

What HSG38 doesn't cover

Comfort and performance

How much light a task needs to be done well depends on how fine the detail is, the age of the person doing it, and how fast and accurately it has to be done. HSG38 leaves those performance levels to industry lighting guidance. Its own table is about safety.

Escape lighting

For the lighting that lets people leave a building when the normal lighting fails, HSG38 refers to a separate code of practice. This page gives no figures for escape lighting.

Differences between neighbouring areas

HSG38 also recommends maximum ratios between the lighting of a work area and the areas next to it, to avoid sudden changes. Those compare two areas rather than points within one, and Luxwork does not evaluate them.

How to measure an area against HSG38

  1. Pick the row. Choose it by the activity. For an area in the first row where someone has to see detail to recognise a hazard, note (a) says to use the row for work requiring perception of detail.
  2. Use a calibrated lux meter. Note its make, model, serial number and calibration certificate. A phone's own light sensor is not a calibrated instrument.
  3. Record the conditions. Whether the lamps had run up to full output, whether daylight was present, and whether every fitting was working.
  4. Decide the measurement plane. The floor, or the surface the work is done on. Record which.
  5. Take readings across the whole area, at points spread evenly over it, corners and edges included. Keep yourself and your shadow off the sensor.
  6. Work out two numbers. The average of all the readings, and the lowest single reading. Compare the average with the row's average figure, and the lowest reading with its minimum figure. Each one is at or above HSG38's figure, or below it.

The whole method is in How to do a workplace lighting survey. For offices, warehouses, stairs and other common areas, Lux levels by area shows which row usually applies.

Recording it so someone else can check

A survey is often read by someone who was not there. For them to check it, the record needs:

A specimen report is set out in What a lighting survey report should contain.

Doing this with Luxwork

Luxwork, a lighting survey app for iPhone and iPad, ships with HSG38's figures. Choose them before you start, and the report states both requirements and where the survey came in against each. It does not stamp pass or fail.

Luxwork launches in autumn 2026

Join the beta list

Test it before launch, on iPhone or iPad. We'll invite testers in small groups through Apple's TestFlight.

Source and licence

Table and figures from HSG38, Lighting at work, second edition, Health and Safety Executive, 1997. © Crown copyright. This page is not published or endorsed by HSE.

Contains public sector information published by the Health and Safety Executive and licensed under the Open Government Licence.