They are not alternatives — they act on different heat and air paths. Draught proofing stops air moving through the gaps a sliding sash needs in order to slide; in Historic Scotland's measured testing it cut air infiltration by up to 86% but moved the window's U-value only from 4.5 to 4.2 W/m²K, a difference the researchers describe as not statistically significant. Secondary glazing does the opposite job: it cut heat loss through the glazing by 63%, taking the U-value below 2.0 W/m²K, and a sealed unit cut air leakage by 97%. Cold glass and traffic noise point to secondary glazing. Draughts, rattle and dust point to draught proofing.
The short answer
A box sash window loses heat two ways. Air blows through the running clearances around the sashes, and heat conducts out through the glass. Historic England's guidance on draught proofing windows and doors states that less than a quarter of the heat lost through a typical traditional window escapes by conduction through the glass — the rest goes as draughts. So on most windows the air path is the larger of the two, which is why draught proofing is the cheaper measure that people notice first.
But that split describes the whole window. Look only at the glazed area and the picture inverts. Historic Scotland's Technical Paper 1 notes that although the glazed area was 55% of the total window area on the window tested, approximately 72% of the heat lost through the window went through the glazing. Draught proofing cannot touch that share, because it does not warm the glass. Secondary glazing is aimed squarely at it.
So the honest comparison is not "which is better". It is: which of the two problems do you actually have? If you can feel air moving, hear the sashes rattle in the wind, or find dust on the cill, that is the air path, and draught proofing is the measure. If the room is cold within a metre of the window even on a still day, or you can hear the street clearly through the glass, that is the glazing, and no amount of brush pile will change it.
What the measurements show
The single most useful piece of evidence in this area is Thermal Performance of Traditional Windows, Historic Scotland Technical Paper 1, by Paul Baker of Glasgow Caledonian University, published October 2008 with corrections in September 2010. It measured a traditional single-glazed sash and case window in a guarded hot box, before and after a professional draught-proofing installation, and then measured the effect of a range of other measures fitted to the same window.
On draught proofing, the paper found air flow at N50/20 fell from 3.5 m³/h to 0.5 m³/h — an 86% reduction in air leakage. On the same window, over the same tests, the standardised thermal transmittance moved from 4.5 W/m²K to 4.2 W/m²K, which the paper states is not statistically significantly different given an overall measurement uncertainty of ± 5.5%, with an average value of 4.4 W/m²K. Historic England cites the 86% figure in its own guidance as the reduction achievable on a sash window in good condition, alongside a general range of 33–50% for draught proofing windows.
On secondary glazing, the same programme measured a 63% reduction in heat loss through the glazing, bringing the U-value below 2.0 W/m²K, from a bare single-glazed centre-of-pane value of 5.4 W/m²K. Sealed secondary glazing also cut air leakage by 97% against the window as received, to an N50/20 figure of 0.1 m³/h.
Read side by side: draught proofing is a very effective air-leakage measure and a negligible thermal-transmittance measure, while secondary glazing is strong on both counts.
How every option compares
Technical Paper 1 did not stop at secondary glazing. It measured the common internal treatments as well, and Historic Environment Scotland restates the headline figures publicly. All of these are reductions in heat loss through the glazing, measured on the same window — they are not reductions in a household's total heat loss, and they should never be presented as such.
| Measure | Reduction in heat loss through the glazing | Resulting U-value (W/m²K) |
|---|---|---|
| Bare single glazing (centre of pane) | — | 5.4 |
| Heavy curtains | 14% | 3.2 |
| Modern roller blind | 22% | — |
| Victorian roller blind | 28% | — |
| Timber shutters | 51% | — |
| Insulated shutters (aerogel blanket) | 60%, and up to 80% if the insulated area is maximised | ≈0.7 at the maximum |
| Secondary glazing | 63% | below 2.0 |
| Combinations (blinds, shutters and curtains together) | Greatest reductions; more than 75% with secondary glazing added | below 2.0 |
| Draught proofing | No significant change (4.5 → 4.2 W/m²K); 86% reduction in air infiltration | 4.2, within measurement uncertainty of 4.5 |
Two things stand out. First, existing timber shutters — which a great many Georgian and early Victorian houses still have folded into the reveals, painted shut — measured 51%, close to secondary glazing, at the cost of easing and repainting them. Second, the combinations row is the highest performer of the lot: Historic Environment Scotland puts secondary glazing combined with blinds and shutters at more than 75%. If a room is genuinely cold, the answer is rarely one product.
Cost
Draught proofing is the better-documented of the two. A full overhaul-and-draught-proof — sashes out, all four sash cords renewed, sashes eased, new parting bead and staff bead carrying brush pile in a carrier, pile routed into the meeting rails — clusters around £250–£400 per window nationally and £328–£588 in London for a standard size, on published specialist price lists. Quotes of £80–£150 a window are for fitting a stick-on strip without any overhaul, which is a different job. Our cost page breaks the figures down by scope and region.
We are not going to publish a per-opening price for secondary glazing, because we do not have a source for one we would stand behind. It is priced by specialist suppliers per opening and varies enormously with frame type, glass specification and whether the units open. Get two or three quotes. As a related reference point, double glazing an existing sash — a different measure again, and one that forces the sashes to be re-balanced because the glass is heavier — is quoted at £650–£1,300 a window by a national trade directory.
On VAT, note that HMRC VAT Notice 708/6 puts draught stripping on the list of energy-saving materials qualifying for the zero rate on installation in residential accommodation from 1 May 2023 to 31 March 2027, after which it reverts to the reduced rate of 5%. The relief covers the installation service and materials supplied by the installer; materials bought by a homeowner are standard-rated. Where draught proofing sits alongside other work, the single-supply test in the notice decides the rate, so ask your installer to confirm rather than assuming.
Appearance and reversibility
Draught proofing is close to invisible once done. The brush pile sits in a carrier let into the parting bead, the staff bead and a routed groove in the meeting rail; from inside the room you see painted beads exactly as before. The trade-off is that it is not fully reversible: the original beads are usually replaced rather than reused, and grooves are routed into the sash timber. That is minor loss of historic fabric, but it is loss, and it is the reason the consent position matters on a listed building.
Secondary glazing is the reverse. It is reversible in principle — an independent frame inside the reveal, removable without altering the sash — but it is visible, and on a fine window with slim glazing bars a second set of sightlines is noticeable. How well a supplier hides the frame members behind existing sightlines is the main thing to judge between quotes. It also has to be planned around shutters: secondary glazing that prevents original shutters from closing removes a measure that tested at 51% on its own.
Listed buildings and conservation areas
This is where the two measures diverge most sharply, and where the received wisdom is usually backwards.
Historic England advises that installing reversible secondary glazing does not generally require listed building consent. The exceptions are set out on pp.34–35 of its Listed Building Consent advice note (published 21 June 2021) and turn on whether the installation damages the window frame, panelling, shutters or other features — so a scheme that screws into original joinery or blocks shutters is a different case from one that does not.
Draught proofing is more nuanced than most sites admit. Historic England states that adding simple self-adhesive draught-proofing strips does not require listed building consent, but that more invasive methods of draught proofing, such as grooving, may affect special interest and may require listed building consent. Routing grooves is exactly what professional draught proofing involves. So on a listed building the low-tech, obviously-reversible option is the one that is clear of consent, and the professional option may not be. Our page on listed buildings and conservation areas sets this out in full, including the position in conservation areas and in Scotland.
In every case, confirm with your local planning authority's conservation officer before work starts, and ask for the answer in writing. Nothing on this page is a guarantee about what any particular authority will decide.
Noise
Noise is the question that sends most people to this comparison, and here the guidance is unambiguous. Historic England's advice on draught proofing windows and doors states that for noise reduction the important criteria are that the windows are well fitted and draught-proofed, but that secondary glazing, with its larger gap — ideally 100mm — between the panes, is a better sound insulator, and that shutters and heavy curtains also work well.
That 100mm gap is the whole mechanism, and it is why secondary glazing fitted tight against the sash performs worse than secondary glazing set back into the reveal. If traffic noise is the reason you are reading this, the depth of the reveal you have available is the specification detail to focus on.
Draught proofing does reduce noise, because sealing the air paths removes the direct route sound was taking, and because it stops the sashes rattling — which on a busy road is often what people are actually hearing. One manufacturer publishes figures of up to 6–10 dB(A) for their own perimeter sealing system, stated as independently tested; treat that as a manufacturer's figure for one system rather than an industry rule. Specialists themselves concede that on a busy road, or with noisy neighbours, draught proofing alone is unlikely to be fully effective. Be wary of any company giving you a percentage noise reduction: noise is measured in decibels, and a percentage figure means nothing.
Maintenance and lifespan
Brush pile is a consumable. Specialists commonly guarantee it for around five years and expect roughly ten years of service in practice, after which the pile slides out of its carrier and is replaced without touching the timber — provided a carrier system was used rather than pile bedded straight into the moulding. The commonest cause of early failure is a decorator painting over the pile; the fix is fresh pile, not a new installation.
Secondary glazing has very little to go wrong — a frame and a pane, with runners or hinges if it opens — but it adds a second set of surfaces to clean, and opening the original sash becomes a two-stage operation. On a window you use daily in summer that is a real everyday cost which shows up in no test report. Mime the movement in the room before committing.
Condensation and ventilation
Neither measure eliminates condensation, and any company that says otherwise is overselling. Condensation forms on the coldest surface in the room. Draught proofing reduces air change without warming the glass, so it can make surface condensation on single glazing worse rather than better. Secondary glazing does warm the room-side surface, which usually helps at that location — but it can move the problem to the outer pane if the cavity is not managed.
Historic England is clear that ventilation has to be maintained in a building that has been effectively draught-proofed: rooms used for living and sleeping need about 0.4 air changes per hour, and between 0.4 and 0.8 ac/h is recommended for older buildings in reasonable condition. Water vapour may need removing at source from kitchens, bathrooms and laundries, and specialist advice should be sought before sealing any room containing a gas or oil burning appliance.
How to decide
A workable decision rule, in the order most households should apply it:
- Repairs first, always. Historic England states that windows should be assessed for repairs before any draught-proofing measures, and that straightforward repairs alone can reduce air infiltration and heat loss by up to a third. Rot at the cill, a dropped sash, failed putty and seized axle pulleys all need dealing with before anything is sealed or fitted in front of the window.
- Draughts, rattle, dust, or a window that no longer opens? Draught proofing. It is the measure aimed at that problem, and rattle in particular it fixes mechanically rather than partially — see rattling sash windows and draughty sash windows.
- Cold glass, or traffic noise? Secondary glazing. This is the measure the data supports for both, and draught proofing will disappoint you if that is what you are buying it for.
- Shutters already in the reveals? Ease and repair them before spending on anything else. At 51% they are the cheapest large improvement available in most period houses, and they are already yours.
- All of the above? Do both, in that order, and design the secondary glazing so the shutters still close.
One last framing point. Draught proofing a house full of box sashes at several hundred pounds a window will not pay for itself on energy alone in any sensible timeframe — the Energy Saving Trust's current figure of about £55 a year is for draught proofing a whole home's windows, doors and floors, not for a sash overhaul. It is bought for comfort, quiet, rattle, dust and keeping original windows working. Secondary glazing is bought for cold glass and noise. Judged on what each is actually for, both are good value; judged on a payback calculation, neither looks clever. See what the draught proofing service includes, or go back to the draught proofing overview.