Almost every bad front door is a good colour in the wrong place. Someone saves a deep green door from a photograph of a Georgian townhouse, paints it onto a 1990s render-and-brick semi, and cannot work out why it looks like a mistake.
The door was never the problem. The house it came from was doing half the work.
The door is an accent, and accents need permission
Your front door is roughly 2% of what you see from the pavement. It gets a wildly disproportionate share of attention because it is at eye level, it is the thing people walk towards, and it is the only part of the facade that moves.
That combination is why a door colour cannot be chosen in isolation. An accent works when it is the one loud thing against a settled background. If the walls are already saying something, the door has to say less — or nothing at all.
So the order is: establish the walls and the roof, then ask what the door is allowed to be.
Three decisions are already made before you start
Before you look at a single door colour, write down what is fixed:
The roof. Covered in why the roof decides your facade — it is the largest surface you are not changing, and it sets the temperature of the whole house. A warm brown roof rules out cool door colours that would otherwise be fine.
The wall colour. The door either contrasts with it or disappears into it. Both are valid; drifting halfway between is not.
The hardware and windows. Black window frames, white uPVC and brown timber are three different houses. A brass letterplate against a cool slate door is a quiet clash that most people feel and few people name.
What works against what
The useful question is not "is navy a good door colour" but "what does this wall colour permit".
Against white or off-white walls, almost anything works, which is why white-walled houses are the ones in all the inspiration photos. Black, deep green, navy, oxblood, mustard. This is the most permissive starting point you can have.
Against grey or greige walls, go either much darker or genuinely saturated. Black, charcoal, forest green, deep teal. Mid-tone colours — sage, dusty blue, taupe — sink into grey walls and read as indecision rather than restraint.
Against warm cream or sand walls, stay warm. Oxblood, terracotta, olive, warm black. A cool blue door against a cream wall is the same mismatch as a cool wall under a brown roof: technically two nice colours, visibly wrong together.
Against brick, treat the brick as the accent and let the door be quiet. Brick is already a pattern, a texture and three colours at once. Black, charcoal, deep green or a white that matches the trim. Competing with brick is a fight you lose.
Against dark walls, reverse the logic entirely. The door now needs to be lighter or warmer, not louder — a natural timber, a warm white, a soft ochre.
Sheen matters more here than anywhere else
On a wall, sheen is a maintenance decision. On a door, it is a colour decision.
The same green in matte, satin and gloss reads as three different colours. Gloss reflects the sky, which means a glossy door takes on the weather: brighter and bluer on a clear day, flat and dull under cloud. Matte holds its colour but shows every fingerprint around the handle.
Satin is the default for a reason. Choose gloss deliberately — on a period property where it is traditional, or where you want the door to be the brightest object on the facade — not because it was what the tin had.
Test it at the door, not at the kitchen table
A door is almost always in partial shade. It sits under a porch, in a recess, behind a storm canopy, or simply on a wall that the sun reaches for two hours a day. That shade is not a minor adjustment — it is the condition the colour will live in permanently.
This is the same problem described in why exterior paint colours never look like the swatch, concentrated into one square metre. A sample held flat in bright daylight is being tested in conditions the door will never experience.
If you are sampling physically: paint a large piece of card, hold it in the actual door opening, and look at it at the three times you actually arrive home.
Then stop sampling and look at the house
The limit of physical testing is that you can only ever judge the door against the door. You cannot see what navy does to the whole facade until the whole facade is navy-adjacent, and by then you have bought the paint.
Photograph the front of the house from across the street, where people actually see it, and try the shortlist on that photograph before you commit to anything. Four candidate colours is ten minutes of looking and one decision, instead of a weekend and a tin you will not use again.
Stop imagining it. Look at it.
One photo of your house, and Exterior Snap AI shows you the finished result in about ten seconds.
