Short answer
UK guidance measures it as a difference in light reflectance value (LRV), not as a difference in hue: Approved Document M treats surfaces as visually contrasting when their LRVs differ by more than 30 points, accepts a minimum of 20 points where illuminance on the surfaces is above 200 lux, and 15 points for door furniture that stands proud of the door. Two strongly different hues with similar LRVs can fail, because people with low vision rely on lightness far more than on colour.
Many people with low vision, and many older people, lose the ability to separate colours well before they lose the ability to see light and dark. A red door in a green wall can look vivid to a designer and indistinguishable to someone with reduced acuity or a colour-vision deficiency if the two are equally light. LRV captures the property that survives: how much of the incident visible light a surface reflects, on a scale from 0 for a perfect black to 100 for a perfect white. Guidance therefore asks for a difference in LRV between surfaces that need to be told apart — door and wall, handrail and wall, step nosing and tread, sanitary fittings and their background.
The figure most often quoted in the UK comes from Approved Document M, the statutory guidance on access to buildings in England. It describes visual contrast as a difference in LRV of more than 30 points between two surfaces. Where illuminance on the surfaces exceeds 200 lux, a minimum difference of 20 points is accepted, because light helps. Door furniture that projects from the door face, and so creates its own shadow and edge, is considered adequate with 15 points. The same guidance refers readers to BS 8300-2:2018 and to research guidance on colour, contrast and perception from the University of Reading. The numbers are thresholds in guidance, not properties of vision; people vary, and more contrast generally helps.
LRV is measured on physical samples, typically to BS 8493. A value quoted for a paint colour describes that paint as tested; the same colour in a different sheen or on a different substrate can measure differently.
An LRV difference is an absolute measure, while screen accessibility uses a ratio of luminances. They behave differently. A 30-point difference between surfaces of LRV 85 and 55 is a luminance ratio of only about 1.5:1, while the same 30 points between 35 and 5 is 4:1. That is one reason the building rule and the web rule are not interchangeable: the building rule gives the same credit to contrasts among light surfaces, which are common in interiors, as to contrasts among dark ones. Neither measure includes glare, texture, pattern or how light falls, all of which change how contrasting two surfaces look in a real corridor.
The common failures are not dramatic. A feature wall creates contrast in the wrong place, drawing the eye away from the door. Shiny floors reflect windows and lights, so a floor with a good LRV difference from the wall still presents a bright streak that reads as a step or as water. Patterned carpets with strong light and dark elements can look like changes of level. A contrast specification that lists LRVs but not finish, lighting and viewing direction leaves most of these problems unaddressed. This page describes guidance; it is not a compliance assessment.
| Example pair | LRVs | LRV difference (points) | Luminance ratio (with flare allowance) | More than 30 points? |
|---|---|---|---|---|
| White wall, mid-tone door | 85 / 55 | 30 | 1.5:1 | Only above 200 lux |
| Light floor, darker skirting | 60 / 30 | 30 | 1.9:1 | Only above 200 lux |
| Mid-tone wall, dark handrail | 45 / 15 | 30 | 2.5:1 | Only above 200 lux |
| Dark wall, near-black fitting | 35 / 5 | 30 | 4.0:1 | Only above 200 lux |
| Pale wall, pale door (fails) | 80 / 60 | 20 | 1.3:1 | Only above 200 lux |
Why: The two colours differ strongly in hue but little in LRV.
Fix: Compare published LRVs (or measure them) and change one surface so the difference exceeds 30 points.
Why: Specular reflections of windows and luminaires create bright patches that LRV does not describe.
Fix: Specify matt or low-sheen floor finishes in circulation routes and check views towards windows.
Why: LRVs were taken for a different sheen, or the lighting is lower than assumed.
Fix: Use LRVs for the actual product and finish, and check illuminance where the 20-point allowance is relied on.
Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.
StandardStrong evidence
Approved Document M Volume 2 describes surfaces as visually contrasting when their light reflectance values differ by more than 30 points, with a minimum of 20 points where illuminance on the surfaces exceeds 200 lux and 15 points for door furniture that projects from the door face.
StandardStrong evidence
Approved Document M defines light reflectance value as the total quantity of visible light reflected by a surface at all wavelengths and directions when illuminated by a light source.
FactModerate evidence
BS 8300-2:2018 is the current British Standard code of practice for accessible and inclusive buildings, and BS 8493:2008+A1:2010 is the current test method for measuring LRV.
Caveat: A draft revision of BS 8493 was issued for public comment in July 2026, so the test method's edition is likely to change.
Source: European Standards online catalogue (EN, ISO, IEC, BSI, ASTM listings)
Colourwise analysisModerate evidence
A fixed 30-point LRV difference corresponds to luminance ratios ranging from about 1.5:1 between light surfaces to 4:1 between dark ones.
Based on: Colourwise calculation treating LRV as a luminance factor in percent and applying a WCAG-style flare allowance of 5 points; see the table on this page.
Caveat: An analogy between two measurement conventions; neither guidance document makes this comparison.
Source: Approved Document M, Volume 2: Buildings other than dwellings (2015 edition incorporating 2020 and 2024 amendments); Web Content Accessibility Guidelines (WCAG) 2.2
Reviewed 29 September 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.