Condensation on windows, and why the glass is the gauge not the fault
Your window is the coldest surface in the room, so it is where the air shows its hand first. Which glazing actually sits below the dew point of a normal room, where the water goes once you have wiped it off, and how many litres a day you are putting into the air without noticing.
By The Damp GuyLast updated
The short version
The water on the glass is a reading, not a fault. Glass is the coldest surface in most rooms, so it is where the air shows its hand first. In a room at 20°C and 60% humidity the dew point is 12.0°C, and anything colder than that runs wet. On a 5°C night a single-glazed window in that room works out at about 10.6°C, so it streams. A modern sealed unit works out at about 16.5°C, so it stays clear. The moisture does not leave with the glazing van. It goes to whatever is now the coldest surface, which is usually the reveal, the window head or the wall behind the wardrobe.
Your window is the coldest surface in the room
Which is why it wets before anything else does, and why it tells you something no other surface in the house will.
Air holds water vapour, and every temperature has a limit to how much it can hold. Cool that air and the limit falls. The temperature at which it runs out of room is the dew point, and any surface below it collects water. BS 5250:2011 puts it plainly: warm humid air cooled to its dew point by contact with cold, non-absorbent surfaces "such as window glass, un-insulated pipework and cisterns" will deposit moisture on them.
Note where glass sits in that sentence. It is one example of a class, not a defect. It wets first because it is the worst-insulated square metre in the room by a distance, and that is exactly what makes it useful. Every other cold surface in the house is hidden behind furniture, plaster or a wardrobe. The glass is the one the house shows you.
The mould threshold sits above the dew point, not at it. BS EN ISO 13788 puts the risk of mould growth at a surface relative humidity above 80%, which a surface reaches while it is still several degrees warmer than the dew point. In the same 20°C room at 60% humidity that line falls at 15.4°C, against a dew point of 12.0°C. BS 5250 reaches the same pair of temperatures from a psychrometric chart and draws the conclusion out loud: designing to avoid mould growth is more onerous than designing to avoid condensation.
Which glazing sits below those lines
For a room at 20°C and 60% humidity on a 5°C night. Read down to whatever is in your walls, and the last column tells you which of the two lines your glass sits on.
| Glazing | U-value (W/m²K) | Inside face at 20°C in, 5°C out | Where that sits |
|---|---|---|---|
| Single glazing in an old metal frame | 5.7 | 8.9°C | Below the dew point. It runs with water. |
| Single glazing in a wood or PVC-U frame | 4.8 | 10.6°C | Below the dew point. It runs with water. |
| Double glazing, uncoated and air filled, 12mm gap | 2.8 | 14.5°C | Dry to the touch, and inside the mould band. |
| Double glazing, soft-coat low-E and argon filled, 16mm gap | 1.8 | 16.5°C | Clear of both lines. |
| The standard a replacement window has to meet today | 1.4 | 17.3°C | Clear of both lines. |
The row that matters is the third one. An uncoated, air-filled sealed unit is the standard specification from before low-E coatings and argon became normal, and it clears the dew point comfortably. It does not clear the mould line. So it never visibly runs, it feels dry to the hand, and it sits in the band where spores germinate anyway. That is the gap between "my windows are fine" and black spots round the frame by February.
The top two rows explain the thing older houses know by heart. A single-glazed window in a heated room is below the dew point at any humidity a person would willingly live in. You cannot ventilate your way out of that. It is the only line on the table where the glass itself is the binding constraint rather than the air.
U-values from SAP 10.2 Table 6e, published by BRE on behalf of the department, for a wood or PVC-U frame. The last row is the standard Approved Document L Table 4.2 sets for a replacement window in an existing dwelling, 1.4 W/m²K or Window Energy Rating Band B. These are whole-window figures covering frame and glass together, so each one is an average: the edge of the glass at the spacer bar and the frame itself run colder than the number shown. The surface temperature is worked out from the U-value and the internal surface resistance of 0.13 m²K/W that BS EN ISO 13788 Table 2 sets for windows and doors. That rearrangement is arithmetic over two published quantities, not a formula any standard prints in one line, and it describes a stated glazing type at stated conditions rather than your window.
Somebody has actually measured this
Glasgow Caledonian University put windows in a hot box for Historic Scotland, at 22°C on the warm side and 2°C on the cold side. The centre of a single-glazed pane measured 12°C. A double-glazed unit measured 18°C. Secondary glazing over the original single pane measured 19°C, which is the finding people with sash windows tend to want.
Those measurements come out warmer than the arithmetic above gives for the same conditions, and that is worth saying rather than hiding. The surface resistance the standard sets is the cautious case, and a sealed test chamber is not a bedroom with a radiator under the window. Both point the same way on the only question that matters here: single glazing in a heated room sits on or under the dew point, and a sealed unit does not. The tested double unit was a slim-profile krypton-xenon one rather than an off-the-shelf sealed unit, and its U-values are centre-of-pane.
Three places the water sits, and three different meanings
They look almost identical through a wet pane at seven in the morning. Only one of them is a fault with the window.
Before anything else, work out which face the water is on. Put a finger on it. If it wipes off from inside the room, it is condensation from your air. If it wipes off from outside, it is condensation from the outdoor air. If it wipes off from neither, it is inside the sealed unit and no amount of ventilation will touch it.
| Where the water is | What it is telling you | What changes it |
|---|---|---|
| On the room side of the glass | The air in the room is above the dew point of the coldest surface in it. The glass is reporting the room, not failing. | Less water going into the air, or a warmer surface. Nothing you apply to the glass itself. |
| Between the panes | The seal has gone. The Glass and Glazing Federation states that condensation in the cavity of a sealed unit "denotes a failure of the unit". | A new glass unit, which can usually go into the frame you already have. |
| On the outside face, early on a clear morning | The unit is insulating well. The outer pane is radiating heat to the sky and is not being warmed from inside to make it up. | Nothing. It clears as the morning warms, and it is evidence the glazing is doing its job. |
The outside-face one catches people out every autumn, and the industry has been unusually straight about it. The Glass and Glazing Federation calls it "a natural phenomenon and a clear indication that the window or door is preventing heat loss from your house". Pilkington, who sell a glass marketed as the answer to it, still say on the same page that it "is not a fault in the glass or window" and that the better the insulation, the colder the outer pane and the likelier it gets. Single glazing almost never does it, because so much heat is pouring through the pane that the outside face stays warm.
Why some windows and not others comes down to line of sight to the sky. A pane looking at open sky on a still, clear night loses heat to it and drops a couple of degrees below the air around it. A pane under a deep eaves, behind a tree or facing a neighbour's gable does not. The GGF says the same in fewer words: it appears on some windows and not others "due to variable micro-climates in differing locations".
Water between the panes is the one case where the glass genuinely is the fault, and it has its own answer, because a failed sealed unit is a replacement job rather than a damp problem.
Wiping the glass does not take the water out of the house
Wiping is worth doing, and it is not a fix. Where the water goes instead is what decides whether anything changes.
Wipe the pane. Genuinely, do it. BS 5250:2011 says condensation on an exposed surface "such as a window pane, where it is easy to see, it should be wiped up immediately or it will encourage the growth of mould", and water sitting in a timber frame will rot it. But be clear about what wiping achieves. It removes water from the glass. It removes none from the house.
The water arrived on the glass out of the room's air. Take it off and the room is very slightly drier for a few minutes, then the air re-equilibrates from everything else evaporating in the house, and the deposition carries on. It carries on at the glass, and it carries on at every other surface that is under the same threshold. The glass was never the only one. It was the first one, and the only one you can see.
The standard names the other ones. Thermal bridges "such as those around doors and windows produce lower local temperatures", which is the reveal and the lintel over your head. Government guidance names "window frames, corners and low points on walls behind sofas or wardrobes". And BS 5250 makes the point that ought to be on the front of every leaflet about this: condensation can also occur out of sight inside floors, walls and roofs, where it is called interstitial, and is "potentially much more damaging".
So treat the wiping as reading a gauge. The useful question is not how to wipe it faster. It is what the fact that you are wiping every morning says about the rest of the room. BS 5250 gives a marker for that too: above 70% relative humidity in the room, the surface humidity of an external element is likely to pass 80%, and if that lasts more than two or three days mould is likely to develop. How cold a particular wall actually runs depends on how it is built, so that is a signal rather than a rule about your house.
Why it is a winter problem, and a bedroom problem
Winter widens the gap at both ends at once. The glass gets colder, because the outside is colder. The air gets wetter, because the windows are shut, the washing comes indoors and nobody is ventilating a house they are paying to heat. Two variables moving the wrong way together is why a window that was fine in October streams in January.
A bedroom concentrates the same thing into eight hours. The door is shut, the heating is off, the coldest part of the night arrives while everybody is asleep, and two people breathing put water into a room nobody is ventilating. Then the curtains trap a layer of still, cold air against the glass, which makes the pane colder than the room it is nominally in. That is why the worst pane in most houses is a bedroom window at seven in the morning, and why it is nothing to do with how clean anybody is.
One more thing worth knowing about glass specifically. The same Historic Scotland testing recorded the top of a pane running warmer than the bottom, 9.6°C against 6.6°C in one test, because the cold air sinks down the inside face. So condensation appearing along the bottom of a pane first is the expected pattern, not a sign of a leak at the sill.
Where the water comes from, in litres
Nobody believes the numbers until they see them, because ordinary living does not feel like carrying buckets of water into the house.
BRE Digest 297 put a range on a four-person household. On an ordinary day, with nothing unusual happening, normal living puts 5 to 10 litres a day into the air. Add laundry drying indoors and a flueless heater and the same household reaches 10 to 20 litres. All of that has to leave the building, or it condenses somewhere inside it.
| Where it comes from | Litres a day | Normal day, or on top |
|---|---|---|
| Four people asleep for eight hours | 1 to 2 litres | Inside the normal daily total |
| Two people active for sixteen hours | 1.5 to 3 litres | Inside the normal daily total |
| Cooking | 2 to 4 litres | Inside the normal daily total |
| Bathing, dishwashing and the like | 0.5 to 1 litres | Inside the normal daily total |
| Washing clothes | 0.5 to 1 litres | On top of the normal daily total |
| Drying clothes indoors | 3 to 7.5 litres | On top of the normal daily total |
| A paraffin heater through the evening | 1 to 2 litres | On top of the normal daily total |
Look at the drying-clothes line. On its own it can be more than the rest of the day put together, which is why BS 5250:2011 says flatly that laundry "should never be hung to dry in unheated or unventilated rooms". The same standard adds the one about flueless paraffin and bottled-gas heaters, which "greatly increase the amount of moisture in the air": burning a kilowatt-hour of paraffin in a room releases about 100 grams of water, and a kilowatt-hour of natural gas about 150. An electric heater releases none.
One honest caveat on these figures, because three respectable sources disagree and picking one quietly would be the wrong move. BRE Digest 297 puts drying clothes indoors at 3 to 7.5 litres a day. BS 5250:2011 Table D.6 puts it at 1.5 kilograms a day. CIBSE Guide A, in its 1999 edition, puts it at 5 to 14 kilograms a day. That is a factor of nine between two of them, on the single line that matters most to a reader with a clothes horse in the spare room. I have published the BRE range because it is the one attributed and readable at source, and named the other two rather than averaging three positions into one figure nobody holds.
There are only two levers
Every product and every tip on this subject is one of two things in disguise: it either takes water out of the air, or it makes a surface warmer. Once you see that, the shopping gets much shorter.
- Make less of it. The table above is your list. Lids on pans, the door shut while the shower runs, and the washing anywhere but a cold bedroom.
- Take it out where you make it. Extracting at the source, at the rates Approved Document F actually specifies, beats spreading moisture through the house and diluting it later. Most bathroom fans that do nothing are undersized rather than broken.
- Dry the air without losing the heat. A dehumidifier pulls the same lever as an open window and keeps the warmth in. It does nothing at all about a cold surface.
- Warm the cold surface. Heating moves that lever and so does insulation, and on a wall the insulation is usually the cheaper of the two over a winter. On the glass itself, the only real move is better glazing, and the table further up tells you what it would buy you.
- Check you are looking at condensation at all. Condensation tracks the cold and penetrating damp tracks the rain. If a patch near a window gets worse in wet weather rather than cold weather, none of this page applies to it.
What you will notice is missing: a spray, a paint, a tub of crystals on the sill. None of them moves either lever at the scale the table above describes. A household putting ten litres a day into the air is not going to be rescued by a pot that holds a few hundred millilitres.
About this resource
The U-values are from SAP 10.2 Table 6e, published by BRE on behalf of the department and used as the default set for existing dwellings where no manufacturer certificate exists. The replacement-window standard is Approved Document L Volume 1, 2021 edition incorporating the 2023 amendments. The surface resistance is the one BS EN ISO 13788 Table 2 sets for windows and doors, which is a different figure from the one it sets for opaque surfaces, and that distinction is the easiest mistake to make here. The dew point and the 80% surface line are already sourced for the condensation pillar, and BS 5250:2011 Annex C arrives at the same two temperatures independently. BS 5250:2011 has been superseded by BS 5250:2021, and it is the 2011 text that is quoted here.
The surface temperatures on this page are worked out, not measured. They describe a stated glazing type at stated conditions, and they are an average across the whole window, so the coldest spots on a real one run colder still. I cannot tell you what your glass is doing and this website does not survey, treat or quote for work.
No costs appear here, for glazing or for anything else. No service life in years for a sealed unit either: BS EN 1279 is a test method and does not state one, and the figures circulating online come from firms selling the replacement. No product is recommended on this page, which is a decision rather than an oversight. The obvious one to sell here is a window vacuum, and the argument above is that the water on the glass is a reading rather than the problem.
If the pattern round your windows tracks the rain rather than the cold, or a patch stays wet through a dry warm spell, that is the point to get an inspection from someone with no interest in selling you the treatment.
Common questions
Does condensation on windows mean the house is too cold?
Not on its own. It means the glass is colder than the dew point of the air in the room, and there are two ways to arrive at that. A warm humid house with one cold window does it. So does a cool dry house with very cold glass. The useful reading is the humidity, not the thermostat. UK professional and public-health guidance settles on roughly 40 to 60% relative humidity in the room, with the HSE allowing up to 70%. Measure that first. Above the range, you have an air problem. Inside it with the glass still streaming, you have a cold-surface problem.
Does condensation on windows mean they need to be replaced?
Only if the water is between the panes. Water on the room side of the glass is the room's air meeting a cold surface, and new glazing raises the surface temperature without changing the amount of water in the air, so the water usually reappears somewhere else in the room. Water between the panes is different, and the Glass and Glazing Federation puts it plainly: condensation inside the cavity of a sealed unit "denotes a failure of the unit". Water on the outside face is not a fault at all.
How much condensation on windows is normal?
There is a published answer, and it is in grams per square metre. BS 5250:2011 gives the quantities: under 30 grams per square metre is a mist that stays put, between 30 and 50 grams per square metre droplets start to run down vertical glass, and above that it runs off sloping surfaces. So the moment you are wiping runs rather than a haze, you have crossed a threshold somebody has actually written down. What it does not tell you is whether your house is fine, because that depends on how long the surfaces stay humid, not on one morning.
How do I stop condensation on windows overnight?
Overnight is the hardest case, because the heating is off, the door is shut and two people asleep put one to two litres into the air of a house over eight hours. There are only two levers. Put less water into the room, by not drying washing in there and by running the bathroom extractor after a shower rather than during it. Or make the glass warmer, which for one night means leaving the door ajar and the curtains open rather than trapping cold air against the pane. Neither is a trick. They are the only two variables there are.
Why is there condensation on the outside of my windows?
Because the unit is working. On a clear, still night the outer pane radiates heat to the sky and drops a couple of degrees below the air around it, and a well-insulating unit is not being warmed from inside to make that up. The Glass and Glazing Federation calls it "a natural phenomenon and a clear indication that the window or door is preventing heat loss from your house". It appears on some windows and not others because of what each one faces, and it clears as the morning warms. Single glazing almost never does it, which tells you everything about why.
Does vinegar stop condensation on windows?
No, and the reason is worth understanding because it rules out most of the other tricks too. Only two things decide whether water forms on your glass: how much water vapour is in the room air, and how cold the glass is. Wiping the pane with vinegar, washing-up liquid or anything else changes neither. What it does is take the water off, which is the same thing a cloth does. If a treatment does not lower the humidity in the room or raise the temperature of the surface, it is not acting on the problem.
Should I worry about condensation on windows?
Worry about what it is telling you, not about the glass. Wipe the pane so the water does not sit in the frame, then treat the wiping as reading a gauge. If you are doing it every morning through the winter, the room is running humid enough that the surfaces you cannot see are collecting too. BS 5250:2011 puts the marker at a room above 70% relative humidity, where the surface humidity of an external element is likely to pass 80%, and says that if that lasts more than two or three days mould is likely to develop. That is the thing to act on.