Best Thermal Curtains 2026: The Number Nobody States

Fifty-one thousand three hundred and thirty-six ratings at 4.5 stars, with 72 per cent five-star and 3 per cent one-star, is the largest body of evidence behind any product I have covered. A curtain is also the kind of product where volume of feedback is unusually informative, because the faults are physical and repeatable: eyelets that pull through, a hem that hangs unevenly, a blackout layer that lets a rim of light around the edges. Fifty-one thousand households have reported on those, and 4.5 stars is what they said.

The listing states what it is selling, which is rarer here than it should be. Two panels, each 46 inches wide and 54 inches long. That is 117 centimetres per panel and 234 centimetres in total, which is enough for a window up to about a metre and a half if you want the folds that make a curtain look like a curtain rather than a blind. The same pair covers a window of 117 centimetres with generous folds and 156 with the minimum a curtain needs, and the arithmetic is the buyer's to do, because the listing does not do it.

The honest note is in the components field. It reads accessories not included, which means the pole is not, the rings are not, and the tie backs are not — and since these are eyelet curtains, a pole is what you need before they can go up. Two other listings in this comparison fill the same field with a tie back or with thread, either of which sounds like something useful and neither of which hangs a curtain. Add 3 per cent of fifty-one thousand ratings to the picture: a very large number of people, an unusually small proportion of them disappointed.

Eighty-one per cent five-star and 1 per cent one-star from 15,002 ratings is the cleanest distribution in this comparison and, at the time of writing, the cleanest of anything I have covered: 12,153 delighted buyers for every 150 disappointed ones. In a category where the product either hangs well or does not, a distribution that narrow means the manufacturing is consistent and the sizing is what people expected.

The reason it is second rather than first is that it does the one thing no other listing does, and it does it in four words: the title gives the size as forty-six by fifty-four inches and then adds, in brackets, width by length. Every other product here quotes two numbers and leaves you to work out which is the drop, and one of them — the Deconovo door curtain at number three — quotes twenty-one different sizes in the same format. Two panels here, so 117 centimetres each and 234 in total, which is a metre and a half of window with folds.

What it does not do is quantify the thermal claim, and neither does anything else on the page. The fabric is described as thermal insulated blackout, and the blackout part is verifiable in the way the curtains behave, while the thermal part has no unit attached to it: no tog, no lining weight, no figure. The eyelet heading is stated and the ring diameter still is not, which matters if your pole is thicker than the standard 28 or 35 millimetres. At £10.18 for two panels with this evidence behind them, it is the pick for anyone who wants the ratio of happy buyers to unhappy ones to be as wide as possible.

This is the listing on the other side of the panel-count problem: a single panel, 52 inches wide and 84 long, which is 132 centimetres by 213, and it says so in the title in the same style as the pairs at number one and two. The difference between a 46 x 54 pair and a 52 x 84 single is not a size step, it is a different job: 213 centimetres of drop reaches the floor from a pole above a standard door, where 137 centimetres of drop stops at the window sill.

The ratings are the second deepest here, 43,602 at 4.5 stars, and the distribution is almost identical to the pair at number one — 71 per cent five-star, 3 per cent one-star — which is what you would expect from the same brand's manufacturing. For a front door with glass, a side door, or a narrow window where a pair would be wasteful, one panel of this width is the right purchase, and the arithmetic is simply that 132 centimetres covers an opening of 66 to 88 centimetres with folds.

There is one trap in the size code and it is the reason this listing is third rather than higher. The same brand and several others sell 52 x 84 as a single and 46 x 54 as a pair, which means the number that looks bigger — fifty-two — is a single curtain, and the number that looks smaller is two. The only defence is the item count field, which reads one here and is correct, and the word panel or pair in the title, which is present on the three best listings in this comparison and absent from the rest. At £9.32 this is a fair price for a door curtain with 43,602 buyers behind it.

Pencil pleat is the older of the two headings in this comparison and the more forgiving one: it gathers into a tight row of pleats along the top, it hangs from a track as easily as from a pole, and it does not require you to own a pole of a specific diameter the way eyelets do. For a bay window, a track, or a house where the hardware is already up, it is the heading to choose, and this is the pair to choose it in: 46 by 54 inches per panel, two panels, 234 centimetres of fabric in total, for £10.44.

The evidence is the third-deepest here: 31,560 ratings at 4.4 stars, with 68 per cent five-star and 3 per cent one-star. That is a slightly weaker distribution than the blackout pair at number two, on a slightly smaller sample, which is why the pencil pleat sits fourth rather than third — the difference between the two products is the heading, and the difference between the two listings is 1 per cent of buyers.

The components field is the familiar disappointment. It reads gathering braid, which is the tape sewn along the top of a pencil pleat curtain and is therefore part of the curtain rather than an accessory, while the hooks you thread into it are not mentioned at all. Two of the listings above fill the same field with a tie back and one with thread. The one thing every buyer of a pencil pleat curtain needs to know before it arrives is that the hooks are a separate purchase, and none of the four pencil pleat listings here says so. Thermal insulated is in the description, as it is everywhere, with no figure attached.

Every other listing in this comparison describes its curtain in inches and leaves it there. This one gives the panel in metres — two metres thirteen long by one metre thirty-two wide, which is the 84 by 52 inches of the title, correctly converted and correctly ordered — and then does something no competitor attempts: it states the weight of the fabric. Seven hundred and fifty grams for one panel. Weight is the closest thing to a thermal measurement in a category where the word thermal has no unit, because a curtain retains heat in proportion to what it weighs and to the air it traps, and a buyer who can compare 750 grams with the weight of another curtain is comparing something real.

It is also the only listing here that labels its own size as width by length in a dedicated field, W 52 inches by L 84, with the word single beside it. That combination — panel count, width label, length label, unit conversion and fabric weight — is five separate pieces of documentation that no other product in this comparison provides more than two of.

The flaw is the claim that goes too far: the title lists soundproof alongside thermal and privacy. A fabric curtain hung on a pole changes the acoustics of a room by absorbing reflections, and a heavy one slightly reduces sound passing through an open gap, but it does not soundproof a door, and no measurement is offered for that either. Twenty-eight thousand five hundred and thirty ratings at 4.5 stars, 72 per cent five-star and 3 per cent one-star, is the third-deepest evidence here, and for a door curtain at £9.99 it is the one whose numbers you can actually check.

Four times the price of the pairs at the top of this page, and the difference is the size and the fabric rather than the finish. At 228 centimetres square, these cover the large window in a living room — the same arithmetic as everywhere else, so 456 centimetres of fabric across the pair for a window up to about three metres, and a drop that reaches the floor from a high pole. For that job the 46 inch pairs are simply not in the running, and a 90 inch square panel is the only size here that is.

The documentation is the reason it sits above the cheaper pairs despite the average. The title gives the size three times — W 90 by L 90 inches, then 228 by 228 centimetres — and it is the only listing in this comparison that converts units for the buyer. That conversion is trivial and it removes the commonest purchase error in the category, which is ordering a drop measured from the window rather than from the pole.

The linen look is a texture rather than a material, and the listing does not pretend otherwise, but it does say blackout in the opacity field, which is the fourth time that field appears on this page and the fourth time thermal appears without a number. Four thousand three hundred and twenty-four ratings at 4.3 stars, with 65 per cent five-star and 6 per cent one-star, is the weakest distribution of the four main listings here: a higher price buys a larger curtain and a heavier look, and in this case it buys slightly less consistent satisfaction than the ten pound pairs. If the window is big, buy it for the size; if it is not, the pair at number one does the job for a quarter of the money.

The second number in a curtain size code is the one that ruins orders, because it is measured from the pole and not from the window, and a 54 inch drop on a pole above a window with a deep sill will stop short of where the buyer expected. Only one listing in this comparison says what the number is, and it says it in the title: 46 by 54 inch drop. The specification then confirms it from the other direction, giving the panel as 137.2 centimetres long by 116.8 wide, which is 54 inches by 46 in the order the title used.

Two panels, so 234 centimetres of fabric, and matching tie backs included — the only listing here where the accessory named in the components field is the one you would actually want, since tie backs are decorative rather than load-bearing. The heading is pencil pleat, so it works on a track or a pole, and the hooks are not in the box, as they never are.

The price of £8.58 makes it the cheapest pencil pleat pair here, and the sample is the reason it sits seventh. Five hundred and seventy-one ratings at 4.2 stars is the thinnest evidence among the mainstream listings in this comparison, and the distribution is the weakest of them: 63 per cent five-star against 7 per cent one-star and 7 per cent two-star, so one buyer in seven gave it two stars or fewer. On a pair this cheap, that is the trade — the best documentation in the comparison at the price of the worst-reviewed product in it.

Eight pounds forty-nine for two panels, which makes this the cheapest pair in the comparison and, at 72 per cent five-star with 5 per cent one-star, not the worst reviewed of them either. That combination — the lowest price and a distribution better than products costing four times as much — is worth stating plainly, because the assumption that a cheap pair is a false economy does not survive this listing.

It is also one of four listings here that declares blackout in a field of its own, which is the honest way to make that particular claim: blackout is about light and it is verifiable when the curtains are up. The thermal claim, which is in the description of nearly every product on this page, is a different property and appears in no field anywhere, which tells you which of the two the sellers can measure.

The weakness is in the components field, and it is the same weakness as three other listings: it names a tie back, which is included and irrelevant, and does not mention the pole that eyelet curtains require to exist at all. A buyer arriving from a search ordered one pair, fitted a pole the same weekend, and the missing bracket is discovered by the person who has to explain it. Two panels of 46 by 54 inches give 234 centimetres of fabric, which is a window up to about a metre and a half with folds, and at this price the arithmetic of buying a second pair for a second window is the easiest decision on the page.

Ninety by ninety inches is the size that turns a pair of curtains into a room divider, and this listing advertises that use rather than hiding it: extra wide, double pieces, ninety by ninety, for a living room, a window or a division of space. Each panel is 228 centimetres wide and 228 long, so a pair carries 456 centimetres of fabric, which is enough for a three metre window with folds or for closing off a doorway. At £36.86 it is priced like the Fusion at number six and for the same reason: this is what large curtains cost.

The documentation is better than most: the title labels the dimensions as W ninety by L ninety, the size is repeated in inches and the pieces are counted. It is one of four listings here that states blackout in an opacity field, and the histogram is strong for the price at 69 per cent five-star and 2 per cent one-star from 6,602 ratings.

The components field is the entry that lets it down: it reads thread. Not hooks, not rings, not a pole — thread, which is either the smallest thing ever included with a curtain or the result of a form being filled in by someone who was thinking about something else. It is a pencil pleat heading, so hooks and a pole or track are what a buyer needs, and a thread is not among them. If the job is a three metre window or a way to divide a room, this is the size to buy, with the caveat that four and a half metres of curtain need a pole rated for the weight.