Condensation in double glazing, and the two versions that are not a fault
Three different things go by this one phrase. Water between the panes is a broken component and has nothing to do with your house. Water on the outside means the glazing is doing its job. Only one of the three is actually damp, and you can tell them apart with a cloth.
By The Damp GuyLast updated
The short version
Wipe the glass. If the water comes off on the cloth, that is ordinary condensation and it is a question about the air in your room. If it wipes off from outside only, the outer pane has gone below the dew point of the outdoor air because so little heat is reaching it, and the Glass and Glazing Federation calls that "strong evidence that heat is not escaping from your house through the glass". If it wipes off from neither side, the water is sealed inside the unit, the desiccant in the spacer bar is saturated, and the unit has failed. That last one is not a damp problem. Your humidity, your heating and your ventilation did not cause it and will not clear it. Only a new sealed unit reaches it.
Wipe the glass. That is the whole diagnosis.
One phrase, three different faults, and two of them are not faults. Where the water sits tells you which you have, and it costs you nothing to check.
People arrive at this question frightened, and usually about the wrong thing. The phrase covers water on the room side of the glass, water sealed between the two panes, and dew on the outside face. Those have different causes, different fixes and different price tags, and only the first of them is damp.
You do not need anybody to tell them apart. Put a cloth on the glass. Which side the water comes off decides everything that follows.
| Where the water is | What it means | What shifts it | A damp problem? |
|---|---|---|---|
| On the room side of the glass. Wipes off with a cloth. | The glass is the coldest surface in the room, and the air in the room met it. | Less moisture in the air, more ventilation, or a warmer surface. | Yes. This is ordinary condensation and the window is the messenger. |
| Between the panes. Wipes off from neither side. | The sealed unit has failed and the desiccant inside the spacer bar is saturated. | A new sealed unit. Nothing else reaches it. | No. Nothing in your room caused it and nothing in your room will clear it. |
| On the outside face. Wipes off from outside only. | So little heat is reaching the outer pane that it has dropped below the outdoor dew point. | Nothing. It goes with the sun or a breeze. | No. It is the glazing working, not failing. |
The middle row is the one that saves people money, and the bottom row is the one that saves them a window. A misted unit will not raise the humidity in your room, will not spread to the unit next to it, and will not grow mould on your wall. It is a failed component behind a pane of glass.
The top row is a different subject, and it has its own page here on condensation on the room side of the glass, because the answer there is about moisture, ventilation and how cold your surfaces run rather than about the window at all. If you are not sure whether what you have tracks the cold or the rain, the timing tells you more than the patch does.
What is actually between the panes, and what broke
The part that failed is inside the bar round the edge of the glass, and you have never seen it. It was losing the argument from the day the unit was made.
A sealed unit is described in BS EN 1279-1 as "an assembly consisting of at least two panes of glass, separated by one or more spacers, hermetically sealed along the periphery, mechanically stable and durable". Pull that apart and there are five things in it that matter.
The spacer is the bar round the edge that holds the panes apart, defined as "a component used to separate the panes and control the width of the gap at the edge of the unit". Inside that bar is the desiccant, "a product designed to lower the water vapour partial pressure inside the insulating glass unit". Round the outside are two seals, not one: the inner sealant touches the cavity, the outer sealant touches the weather, and together they form the edge seal, which the standard says is "designed to ensure that moisture and gas transmissions between the inside and outside of the unit are limited". Note the word. Limited, not prevented.
In practice the inner seal is usually polyisobutylene, which a glass manufacturer describes as "a non-curing component that remains pliable over the lifecycle of the IGU" whose job is to provide "a barrier to gas and water vapor transfer". The outer one is silicone, polyurethane, polysulfide or hot melt butyl, and it is structural: it bonds the glass to the spacer and protects the inner seal from the elements.
That is the whole machine. And the failure is one sentence long, from the same manufacturer: "From the time of fabrication until IGU failure, moisture will slowly penetrate the seal. Desiccant adsorbs this moisture until it reaches saturation. At this point moisture will begin to accumulate in the IGU and fogging between the panes may occur. This is what is referred to as a 'seal failure'."
Read that again, because it settles the question people actually came here with. The desiccant was always absorbing moisture, from the day the unit was made. The day it mists is the day the desiccant ran out of capacity. Nothing happened in your room that morning.
The GGF's own consumer guide puts the same thing as a verdict: "Condensation within the cavity of a hermetically sealed double glazed unit or Insulating Glass Unit (IGU) denotes a failure of the unit", and "condensation will not form on the inside of a correctly functioning Insulating Glass Unit".
One honest qualification, because the GGF's own quality standard gives two routes rather than one. Interstitial condensation happens "either when the unit seal has failed and the cavity is saturated with moisture or when an internal component, e.g. Georgian bars, etc., has deteriorated and given off organic solvents". Both of those are inside the unit. Neither of them is your house.
Three names for it, and two of them are the same word
If somebody quotes a standard at you, check which document they are quoting first. The phrase "external condensation" means two different things depending on which one it came from, and only one of those meanings is the good news.
| What you would call it | GGF Data Sheet 4.10 | BS EN 1279-1 |
|---|---|---|
| Water between the panes. Misted, blown, steamed up. | Interstitial condensation | Internal condensation |
| Wet glass on the room side, worst round the edges. | Peripheral condensation | External condensation |
| Dew on the outside of the glass. | External condensation | External condensation |
BS EN 1279-1 defines external condensation as condensation appearing "either on the room side surface or on the external surface", so in that document the word covers both faces you can touch. The GGF data sheet splits the same ground three ways, and reserves "external" for the outside face alone. Two documents, one phrase, opposite scopes.
Nothing you do in the room causes this, and one thing about the frame does
The clearest attempt to put numbers on what shortens a sealed unit's life is a 2025 Danish service-life model, and its finding on the indoor side is blunt. It scores six indoor-environment variables and gives every one of them a factor of 1.0 even in its worst case. Internal climate, the heating and ventilation system, and solar loading are each annotated "Negligible impact". Room humidity and moisture sources are each annotated "Not within scope". Indoor environment is the only one of the model's seven factors with no downside at all, against 0.86 for how well the work was done and 0.94 for the quality of the components.
The one qualification the same table carries is worth having: sustained high humidity in a wetroom is a risk to the stability of the window frame, and through the frame to the edge seal. That is a slow, indirect chain over years. It is not a consequence of last night's shower.
What does shorten the unit's life is water, sunlight and temperature. The daily temperature swing plus wind load creates "a pumping action on the glass panes, stressing the edge seal, and testing its elastic recovery". And liquid water sitting against the edge is the direct route in: "the secondary sealant will absorb liquid water into the polymer matrix and begin swelling, eventually opening the primary seal".
Which brings you to the one thing on this page that is genuinely about the building rather than the glass. Water sits against the edge seal when the frame does not drain. LABC Warranty requires glazing to have "a drained and ventilated bottom bead with a minimum gap of 5mm between the edge seal of the insulated glass unit and the bottom channel of the frames glazing rebate". Premier Guarantee asks for "a 5mm drainage path between the edge of the IGU and the rebate" all round, and describes the sequence: trapped water freezes, contributing to the seal failing, and "it is not until the spring where misting in the IGUs occurs, can the issue be seen". The Danish model puts the same rule at more than 4mm, and gives the reason in one line: "a water droplet cannot span 4 mm".
Two practical consequences. If the units on one elevation keep going and the ones on another do not, drainage and exposure are worth looking at before anyone blames the glass. And a Pilkington technical manager has said in trade press what the warranty providers imply: "poor installation can significantly reduce lifespans".
Condensation on the outside is the window working
This is the version that looks most alarming and is the only one that needs nothing doing at all. The pane is cold, and it is cold because the unit is keeping your heat indoors.
Dew on the outside face of the glass means the outer pane is colder than the outdoor air is willing to stay wet above. It is cold because heat is not reaching it, and heat is not reaching it because the unit is doing what it was sold to do.
The GGF states the mechanism and the verdict in the same paragraph. Modern units contain low-emissivity glass with "a high-performance transparent coating that reflects heat from radiators or fires back into the room. As a result, the outer pane of glass does not get warmed by heat escaping from inside the building through the glass and remains cooler." Its conclusion is not hedged: "This is strong evidence that heat is not escaping from your house through the glass in your window."
And, in terms, on the question of a defect: "The presence of external condensation is an indication that the glazing is thermally efficient and should not be considered detrimental. The more thermally efficient the glazing, the higher the likelihood of condensation."
The same body's quality standard for glass units says it twice over. External condensation "is not a fault, and can be seen as the product being effective in keeping heat in", and its acceptance criteria for an installed unit include the line that settles any argument with an installer: "condensation, internal or external, is not related to a seal failure". A glass manufacturer agrees, and is worth quoting precisely because it has every incentive not to: on a page inside the product range for its own anti-condensation glass, it says this "is a naturally occurring phenomenon which is more prevalent the more highly insulating your windows are and is not a fault in the glass or window".
- When you will see it. The GGF gives the conditions as "a variable combination of high relative humidity and clear cold conditions commonly experienced in spring and autumn, but can occur at any time of year". Clear, still nights. Gone by mid-morning.
- Why only some windows. "It is often the case that external condensation will appear on some windows but not on others due to variable micro-climates in differing locations." In a house with glazing of mixed ages, that means it turns up on the best units first.
- What to do about it. Nothing, and the GGF says as much: "as this is caused by external atmospheric conditions, little can be done to prevent this condition at certain times of the year". Anybody quoting to cure it is quoting to cure a symptom of the thing working.
Worth holding on to when the same logic turns up indoors. New glazing that runs dry can push the condensation onto your walls instead, because the water goes to whatever surface is now the coldest, and a wall is a worse place for it than a pane you can wipe. Warming that surface is the lever there, not the window.
Repair it, replace it, or leave it
Three real options, and the third one almost never gets offered, because nobody makes any money out of it. Here is what each one actually gets you.
Start with the money, because the shape of it decides most of this. Every published source that prints both a sealed-unit price and a whole-window price puts the unit at a fraction of the window. They disagree about the figures. Not one disagrees about the direction.
| Source, and what it sells | Sealed unit only | Whole window | The gap |
|---|---|---|---|
| MyJobQuotelast updated 11 July 2026, a lead-generation platform, so its incentive is enquiries | £55 to £145 | £350 to £650 | 4 to 6 times |
| ExpertSurepublished September 2026, a comparison site earning on quote requests passed to installers | £80 to £180 | £350 to £600 | 3 to 4 times |
| Window Fittersno article date shown, an installer selling both replacement units and whole windows | £75 to £165 | £450 to £750 | 5 to 6 times |
| Restoration Glazingno publication date shown, a glazing repair company that sells the job it is pricing | £60 to £250 | £350 to £2,400 | 6 to 10 times |
| The Eco Expertsupdated 29 July 2026, a comparison and lead-generation site earning on quote requests | £100 | £1,000 | 10 times |
Every figure is reproduced as its publisher states it, and nothing is averaged: the multiple in the last column is worked inside one source's own two figures, never across two. Two of the sources above are lead-generation platforms whose income is enquiries, one is an installer and one is a repair firm. The two directories ranking at the top of this search both refuse automated access, so nothing is quoted from either of them here.
Two figures the table leaves out, because their publishers print only one side. A national repair franchise puts a standard casement at £130 to £300, and it is the only source read here that discloses a call-out or minimum visit fee, at £50 to £80. That floor is the most useful thing on this page for anyone holding a cheap quote, because it means a £55 unit is not a £55 job and doing every failed unit in one visit is the cheaper route. At the other end, a national installer puts a uPVC casement at £700 to £1,800 including installation, which starts roughly where the cost guides finish. That is not an error in anybody's arithmetic. It is the difference between a local glazier and a surveyed, guaranteed, nationally advertised job.
What a defogging repair actually does to the unit
There is a trade that clears a misted unit in place instead of replacing it, and you will see it advertised. A straight account of it is hard to come by, and that is a problem of incentives rather than of physics. Everyone who dismisses it sells replacement units. The firm that endorses it sells the repair and will not describe its own method beyond the word valves. So take the process description from the firm that sells the kits to the trade, which has no reason to understate what they do.
Two small holes, 6mm, are drilled "either in the glass or through the side spacer bars". A drying agent goes in, then an anti-fog polymer on the internal glass surfaces, then a sealer "to create a seal across the bottom". Then vents go over the drilled holes, and here is the sentence that decides what you are buying: the vents are there to "control the amount of air that enters the unit".
So the unit is not resealed. It is vented, on purpose, and that is the mechanism rather than a side effect. A home services site states the consequence plainly: after treatment the unit is no longer hermetically sealed. Drying can take anywhere from a few hours to, by the same supplier's account, twelve weeks.
The argon question is usually argued badly by both sides. An installer selling replacements says argon "cannot be re-injected or pressurised on site once the hermetic boundary is broken", which is true. The kit supplier concedes the other half: once the seal fails, the heavier gases escape and moist air replaces them. Put those together and the argon was already gone before anyone arrived with a drill. Drilling does not lose you the gas fill. The failure did.
The desiccant is a different matter and the trade does not claim otherwise. No published description of this process includes replacing or regenerating it. A vent takes over the job the desiccant used to do, by swapping the cavity air with outdoor air rather than by drying it.
The one number that tells you whether it will work on your window
This is the fairest thing published about the repair, and it comes from the firm selling the kits rather than from anyone criticising them. It grades the damage in three stages.
| Stage | What you can see | What the trade says it can do |
|---|---|---|
| Stage one | Fog, mist and water. | Removable. |
| Stage two | Marking that has started to set. | Can sometimes be removed, or significantly reduced. |
| Stage three | Baked-on streaks and tide marks. | Cannot be removed. |
The same source adds that the process "will not however remove stains or streaks inside the unit that are baked on or are leaching from the glass", and that afterwards "the damage will not get any worse". The firm selling the service says its process removes "early stage staining", and the qualifier is doing the work in that sentence. An installer on the other side of the argument describes what stage three looks like: mineral deposits baked onto the coating, appearing as milky white streaks or tracks "that never clear, even in direct warm sunshine".
So go and look at your window. Soft haze and droplets is stage one, and every party to this argument agrees that clears. White tramlines is stage three, and the firm that sells the repair agrees it does not.
One hard limit, and it matters more than the money. Toughened safety glass cannot be drilled through the face. An installer's description of what happens if you try is that it "will cause it to instantly shatter into thousands of pieces", and the national repair service excludes toughened units in timber frames from its own process. A homeowner has no reliable way to tell toughened from float glass except by finding the kitemark etched in a corner. That is the reason there is no product recommended anywhere on this page.
What it costs, and who is telling you
- Around £45 plus VAT. Which?, in May 2025, and the only figure here published by somebody selling neither the repair nor the replacement.
- £40 to £80 per pane. A national repair franchise, in August 2026, which sells the competing replacement unit.
- £50 to £100 per window. MyJobQuote, in July 2026, calling it resealing and labelling it short-term.
- £50 to £100 per unit. A home services directory, reviewed in May 2026.
How long it lasts is disputed entirely along commercial lines, which is the most honest thing anyone can tell you about it. The firm selling the repair offers a ten year warranty. A firm selling replacement units says two to five years. Another firm selling replacement units says six to eighteen months. Nobody has published field data, and neither side of this trade has published a checkable test of what it does to the unit's thermal performance either. The only price here from a publisher selling neither outcome is Which?, at around £45 plus VAT in May 2025, with the judgement that "if successful, this is likely to be a short-term fix".
Before you pay anyone, check whether somebody already owes you it
Which? puts typical double glazing warranties at "10 or 20 years, although some are lifetime guarantees", and a Pilkington technical manager put warranty periods at 5 to 20 years. Which? is also blunt about what a misted unit means on a newer window: seals broken or no longer working properly, and "this shouldn't happen with new double glazing". So the first question is the date, not the quote.
Three things get called a guarantee here and they are not the same thing. The manufacturer's guarantee is on the unit. The installer's is on the work. An insurance-backed guarantee is neither of those: FENSA describes it as "an insurance policy that provides cover for a guarantee in the event the company that issued the guarantee ceases to trade". It is not extra cover. It is cover for the company disappearing.
The GGF's consumer code sets no minimum guarantee period at all. Its entire requirement is that a member "must provide the consumer with clear and accurate details of any guarantee, including the period of the guarantee and any conditions attached to it". If you have moved house since the windows went in, the code says an unexpired guarantee can transfer to the new owner on the terms in the original contract, and that the member may inspect first and "may charge a reasonable transfer fee and inspection fee".
Here is the part almost nobody tells you: a misted unit is not a FENSA complaint. FENSA's own complaints service states that it cannot help with "customer care, product quality, compensation, refunds, contractual issues or guarantee". What it covers is whether the installation complied with building regulations, for an approved installer still trading, within the last six years, in England or Wales. On the product itself it points you straight back to the installer's guarantee. If you go to FENSA with a blown unit you will lose a fortnight.
The routes that do work: the installer's own complaints procedure first, because every scheme requires that you exhaust it. The GGF runs a free conciliation service, though only for domestic customers of its full members, escalating to TGAS. And underneath all of it sits the Consumer Rights Act 2015, which the GGF's own code acknowledges when it says a member's guarantee "does not take away or diminish the consumer's statutory rights". Which? advises raising it as soon as you notice, and following the conversation up in writing.
And the option nobody selling windows will offer you. Leaving it alone is a real answer. A misted unit is an appearance problem plus some lost insulation. It will not spread, it will not wet your walls, and the damage stops getting worse once the glass has stained. If you are selling the house, or the frames are near the end anyway, doing nothing until the whole window is due is a decision rather than a failure to act. What decides it is the frames: one comparison site puts it plainly, that if the frame is the problem then replacing the glass "will just be a waste of time and money".
About this resource
The technical positions here come from BS EN 1279-1, the Glass and Glazing Federation's April 2025 consumer guide to condensation and its January 2010 data sheet on the visual quality of insulating glass units, a glass manufacturer's own technical brochure, and a February 2025 Danish service-life study. The drainage requirements are LABC Warranty's and Premier Guarantee's, both structural warranty providers, which is why they carry weight here: they pay the claims and sell no glazing. The prices are every published figure I could reach, each attributed to the source that published it and left in that source's own units, with what that source sells stated beside it. Nothing is averaged.
Four limits worth stating plainly. BS EN 1279 is paywalled: the 2004 Part 1 text was read and the 2018 editions were not, so nothing on this page quotes a test threshold from the standard. The GGF data sheet used here is the January 2010 edition, and a newer version sits behind a member login. The Danish study is funded by window manufacturers and its authors state that parts of it were drafted with the help of generative AI models. And the glass manufacturer quoted on how a unit is built is writing for a North American audience, so it is used for the anatomy and the failure chain and for nothing about UK practice.
Left off on purpose. No trade body position on whether defogging is a legitimate repair: a line to that effect circulates word for word across a dozen commercial sites and it is not on the federation's own website. No thermal performance figure for a defogged unit in either direction, because nobody on either side of that trade has published a checkable test. No duration for an insurance-backed guarantee, because none of the three schemes publishes one. No pounds-per-year figure for living with a failed unit. No product recommendation, deliberately: the only thing you can buy here is a drilling kit, and drilling toughened glass shatters it.
England and Wales for the FENSA scope, and the GGF's note that the insurance-backed guarantee duty does not apply in Scotland or the Republic of Ireland. This site does no glazing work, takes no referral fees and cannot see your window. If the mist is on a unit that might still be inside its guarantee, the cheapest hour you will spend is finding the paperwork before you ring anybody for a price.
Common questions
Can condensation in double glazing be fixed?
Yes, in two different senses, and they leave you holding different things. A new sealed unit puts the glass back to what it was. An in-situ defogging repair drills the unit, dries the cavity and fits vents, so what you get back is a vented unit rather than a sealed one. The firm that supplies the kits says so itself: the vents are fitted to "control the amount of air that enters the unit". Which? put the drilled repair at around £45 plus VAT in May 2025 and said that if successful it is likely to be a short-term fix. Published prices for a replacement unit start around £55 and run past £300, depending on size and on who is quoting.
Will condensation between panes go away?
The water moves about. The fault does not. Guardian Glass sets out the chain plainly: moisture slowly penetrates the seal, the desiccant adsorbs it until it reaches saturation, and after that moisture accumulates in the unit and the panes fog. The desiccant does not dry out again, so nothing reverses it. What you will see is the mist clearing when the sun hits the glass and coming back on the next cold morning, which is exactly what makes people think it is sorting itself out.
How much does it cost to replace a sealed unit in a double glazed window?
Published figures, each as its publisher states it and none of them averaged: £55 to £145 fitted from MyJobQuote (11 July 2026), £80 to £180 from ExpertSure (September 2026), £75 to £165 for a standard casement from an installer, and £130 to £300 for the same casement from a national repair franchise, plus a call-out floor of £50 to £80. Notice who publishes what. The lowest figures come from lead-generation platforms whose income is enquiries. The higher ones come from firms that actually turn up, and only those disclose a minimum visit charge. That charge is why a £55 unit is not a £55 job, and why doing every failed unit in one visit is the cheaper route.
Will a hair dryer get rid of condensation between glass panes?
No, and the question has two different problems folded into it. A hair dryer reaches the room side of the glass and nothing else. A window repair company that sells the repair puts it plainly: waving a hairdryer over the window will dry the condensation, "but this is very much a short term solution". So if your condensation wipes off with a cloth, you do not have a failed unit at all, and you are drying a humidity problem with a very expensive towel. If it does not wipe off from either side, the water is sealed inside the unit and nothing you point at the glass will reach it.
Is it worth replacing 20 year old double glazing?
Twenty years is where the published opinion starts to turn, and it still does not decide it. Which? says double glazing "can become worn and problematic" once it is twenty years old, and puts typical warranties at 10 or 20 years with some lifetime guarantees. A Pilkington technical manager put warranty periods at 5 to 20 years, with actual lifespans often longer. What actually decides it is the frames, not the date. One comparison site makes the point bluntly: if the frame is the problem, replacing the glass "will just be a waste of time and money". Sound frames with a couple of blown units is a glass job. Frames that have had it is a window job.
Is it cheaper to replace glass or the whole window?
Much cheaper, and the size of the gap is the most consistent finding in the whole published cost literature. Every source that prints both figures puts the sealed unit at somewhere between a third and a tenth of the whole window. MyJobQuote: £55 to £145 against £350 to £650. ExpertSure: £80 to £180 against £350 to £600. The Eco Experts: £100 for the pane against £1,000 for a complete replacement. They disagree on the numbers and not one of them disagrees on the direction, which is worth more than any single figure, and it is the fact the commercial pressure on this subject works hardest to blur.