Short answer
Slightly, as far as the best study can tell, and not in the way colour charts suggest. In 153 English primary classrooms, design features together accounted for 16% of the range in a year's progress; colour carried about a ninth of that modelled effect, and rooms with light walls plus some stronger colour did best — neither bare nor saturated throughout. No particular hue was tested. Building requirements say nothing about hue: they set how much light surfaces must reflect and how strongly doors, switches and stair edges must contrast.
Barrett and colleagues at the University of Salford assessed 153 classrooms in 27 primary schools across Blackpool, Hampshire and Ealing, and compared the physical features of each room with the progress its 3,766 pupils made over one school year in reading, writing and mathematics. Multilevel models separated the room from the pupil, allowing for starting attainment, free school meals, special educational needs and English as an additional language. Seven features of the room turned out to matter. Between the most and least effective classrooms the model predicted a difference of 7.93 National Curriculum points for an average pupil, which is 16% of the 50-point range in progress across all pupils. Light, temperature and air quality supplied about half of that; colour supplied 11%.
Taken at face value, colour's share comes to about 0.9 of a National Curriculum point between the best- and worst-scoring rooms. That is Colourwise's arithmetic on the paper's figures, and it is small.
Colour was scored by the researchers for the whole room — walls and how much of them was coloured, floor, blinds, furniture and display — on a scale of brightness, not by hue. Its relation to progress was curved: rooms at either extreme did worse than rooms in the middle. The pattern the authors describe as best is light or white walls with one feature wall or an ordered display in stronger colour, and bright furniture or carpet as accents; strong colour as the main theme of the room scored poorly, and so did a room with none. The study therefore says something about quantity and placement of colour and nothing about whether blue, green or yellow is better. It is also correlational: schools that keep rooms in good order may differ in other ways the model did not capture.
The Department for Education's technical requirements for school and college buildings in England, in the January 2026 Technical Manual, treat colour as reflectance and contrast. Ceilings are to have a light reflectance value of at least 70% and walls at least 50%, unless natural materials are used; floors between 5% and 40%, the upper limit being there to hide scuffing. Door frames and switch plates must differ from the wall by 30 LRV points and stair nosings from the floor by the same, following Approved Document M, which applies to public buildings generally. Interior lighting is to be 3,500–4,000 K with a colour rendering index of at least 80. Noticeboards in a teaching space are capped at a fifth of the wall area where there are no sprinklers — a fire rule, not a view about distraction. Nowhere does the manual name a hue.
These are contract requirements for centrally funded projects, not law for every school, and the manual leaves final colours to be agreed with the school.
Light was the largest single factor in the classroom study, and wall and ceiling reflectance is how colour acts on light. Pale surfaces pass daylight deeper into a room and soften the jump in brightness between a window or a luminaire and what surrounds it; dark ones absorb it. The manual's daylight calculations assume a fifth of each wall will be covered with posters and work at a reflectance of 0.2, which pulls even white walls well down. The table works that through for a few schemes: a single dark feature wall costs about as much average reflectance as repainting the whole room a pale tint, and a mid-tone at the 50% minimum ends up in the low forties once display is counted. Read literally, the 50% minimum applies to every wall finish, so on a project built to the manual a deep feature colour would have to be agreed as a departure. Wherever one is used, it is best kept away from the window wall and the board, so that it does not sit next to the brightest things in view.
Advice that blue aids concentration, green calms or yellow stimulates young children is common in design writing and paint marketing. The classroom study cannot support it, because hue was not one of its variables, and its own review of earlier work lists findings that point in different directions: lower learning scores in a heavily decorated room than a sparse one in one experiment, and less off-task behaviour after off-white walls were repainted in saturated colours in others. Those were small studies of different ages, tasks and settings. The defensible position is the one the healthcare page reaches for hospitals: treat mood claims as convention and preference, ask the people who use the room, and spend the certainty on light, reflectance and contrast. Children's needs differ, and some pupils are more sensitive to visual clutter than others; that is a reason to consult the school, not a finding about colour.
| Parameter | Design principle | Share of the classroom effect | Shape of the relationship |
|---|---|---|---|
| Light | Naturalness | 21% | As entered in the model |
| Temperature | Naturalness | 12% | As entered in the model |
| Air quality | Naturalness | 16% | As entered in the model |
| Ownership | Individualisation | 11% | As entered in the model |
| Flexibility | Individualisation | 17% | As entered in the model |
| Complexity | Level of stimulation | 12% | Curved: a middle level scored best |
| Colour | Level of stimulation | 11% | Curved: a middle level scored best |
| Surface | Light reflectance value | Required contrast | Qualification |
|---|---|---|---|
| Ceiling finishes | At least 70% | None stated | Unless natural materials are used. |
| Wall finishes | At least 50% | 30 points against door frames and switch or socket face plates | Unless natural materials are used; the contrast follows Approved Document M. |
| Floor finishes | 5–40% | 30 points against stair nosings | The upper limit is there to avoid visible scuff marks. |
| Walls as modelled for daylight | No limit set | None stated | Daylight calculations assume a fifth of the wall is covered by posters and similar items at a reflectance of 0.2. |
| Scheme | Painted walls, average LRV | With a fifth of the wall under display | Against the 50% wall minimum |
|---|---|---|---|
| White throughout (LRV 85) | 85 | 72 | Average stays above it |
| Pale tint throughout (LRV 70) | 70 | 60 | Average stays above it |
| Mid-tone at the 50% minimum | 50 | 44 | Falls below once display is counted |
| White with one mid-tone feature wall (LRV 40, a quarter of the wall area) | 74 | 63 | Average stays above it |
| White with one dark feature wall (LRV 12, a quarter of the wall area) | 67 | 57 | Average stays above it |
| Pale tint with one dark feature wall (LRV 12, a quarter of the wall area) | 56 | 48 | Falls below once display is counted |
Why: Average wall reflectance has fallen well below what the daylight design assumed, and display lowers it further.
Fix: Keep most of the wall area pale and confine the strong colour to one wall or to furniture and display.
Why: A dark surface next to bright glazing or a lit board raises the brightness contrast in view.
Fix: Put the feature colour on a wall away from the window and the teaching wall.
Why: Colour-psychology claims are being presented as educational evidence.
Fix: State hue as identity and preference, and set measurable targets for reflectance, contrast and daylight instead.
Why: Frame and wall have similar light reflectance values whatever their hues.
Fix: Check the LRVs and separate frame and wall by 30 points.
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.
FactModerate evidence· England (Blackpool, Hampshire, Ealing)
In a study of 153 classrooms in 27 English primary schools and 3,766 pupils, seven classroom design parameters — light, temperature, air quality, ownership, flexibility, complexity and colour — together accounted for 16% of the variation in pupils' academic progress over a year.
Measured: 3,766 pupils in Years 1–6 in 153 classrooms of 27 primary schools. Method: Researcher assessment of each classroom; multilevel regression of one year's progress in reading, writing and mathematics with pupil-level covariates. The 16% is the modelled gap between the most and least effective classrooms (7.93 National Curriculum points) as a share of the 50-point range in pupil progress.
Caveat: Observational and correlational; primary schools in one country; a single research group's assessment scheme.
FactModerate evidence· England
In that study colour accounted for 11% of the modelled classroom effect and light for 21%; colour's relationship with progress was curvilinear, with an intermediate level best, and light or white walls with a feature wall or bright accents correlated best with progress.
Measured: 3,766 pupils in 153 primary classrooms.
Caveat: Colour was scored on a brightness scale for the whole room; individual hues were not compared, so the result cannot rank colours.
StandardStrong evidence· England
The Department for Education's Technical Manual of January 2026 sets minimum light reflectance values of 70% for ceiling finishes and 50% for wall finishes, unless natural materials are used, and requires floor finishes between 5% and 40%.
Caveat: Employer's requirements for Department for Education construction projects, January 2026 edition; not a statutory duty on every school, and not a compliance assessment.
StandardStrong evidence· England
The same manual requires a 30-point light reflectance value difference between walls and door frames or face plates and between floors and stair nosings, in line with Approved Document M, specifies interior LED lighting of 3,500–4,000 K with a colour rendering index of at least 80, and limits noticeboards in teaching spaces to 20% of wall area unless automatic fire suppression is installed.
Caveat: January 2026 edition; summarised, not quoted.
Source: Design standards: employer's requirements — Technical Manual (technical requirements for the design and construction of school and college buildings, January 2026); Approved Document M, Volume 2: Buildings other than dwellings (2015 edition incorporating 2020 and 2024 amendments)
Colourwise analysisLimited evidence
Colour's 11% share of a 7.93-point modelled gap between the most and least effective classrooms comes to about 0.9 National Curriculum points, under 2% of the 50-point range in pupils' progress.
Based on: Colourwise multiplication of two figures published in the paper: colour's share of the modelled classroom effect and the modelled range between classrooms. The paper does not state the product.
Caveat: A rough scale, not an effect size from the paper; shares of a multilevel model do not divide cleanly, and the estimate inherits the study's correlational design.
Colourwise interpretationLimited evidence
No particular hue has good evidence of improving learning; the supportable design guidance concerns how much strong colour a classroom carries, how much light its surfaces reflect and whether doors, edges and fittings contrast with their backgrounds.
Based on: Colourwise reading of the 2015 classroom study, which scored colour by brightness and quantity and not by hue, together with the Department for Education's requirements, which specify reflectance and contrast and no hue.
Caveat: Not a systematic review. Studies of particular hues exist but are small and inconsistent; a stronger study could change this.
Reviewed 1 October 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.