How to Choose a Waterproof Jacket

Almost every disappointing waterproof jacket was bought the wrong way round: fabric first, use second. The fabric is the last decision, because it is the one most constrained by the others.

Four questions do most of the work. How long will you be in rain without shelter — twenty minutes between buildings, or a full day on a ridge? How hard will you be working, since a body producing a lot of heat and moisture needs venting more than it needs a higher pressure rating? How cold will it be, because being wet is a comfort problem at 15 °C and a safety problem at 2 °C? And what will abrade the garment: a pack, a harness, a climbing rope, or nothing at all?

Write down those four answers before reading a single specification. They tell you whether you are shopping for a light packable shell, a hard-wearing mountain garment, or something in between — and that decision, not the membrane brand, determines whether you will be comfortable.

A hydrostatic-head figure is a laboratory result for a piece of fabric. Water pressure is raised against one face of a specimen until water is forced through in three places; the pressure at that point is the result. It measures a material under steadily rising static pressure, not a garment in wind-driven rain.

Gore, which publishes no such figure, argues the point directly: it describes the water-column test as not relevant to real-life situations, because pressure on a worn garment is not static and a falling raindrop is dynamic. It reports that EMPA, the Swiss federal materials laboratory, found garments made from fabrics rated below 120 mm remained fully waterproof under simulated rainfall. That is a manufacturer arguing against a metric it does not publish, so read it with that interest in mind — but the underlying physics is sound, and it is the clearest available warning against treating the number as a durability-of-dryness score.

Breathability is worse still for comparison shopping. An MVTR value depends entirely on the method used to obtain it, and the standards say so themselves: ASTM E96 states results from its different procedures are not directly comparable, and ISO 15496 states its cup result cannot be used to classify vapour resistance against physiological values. So a jacket advertising 20,000 g/m²/24 h is not demonstrably more breathable than one advertising 15,000 by another method — or than one publishing nothing.

The one genuinely comparable framework is the rain standard for protective clothing. EN ISO 24232:2025 superseded EN 343:2019, whose withdrawal took effect on 22 October 2025. Its waterproof classes require at least 8,000, 13,000 and 20,000 pascals, and the standard itself notes that 1,000 Pa corresponds to about 102 mm of water — so its top class sits near 2,000 mm, far below the figures printed on recreational hangtags. It also classifies evaporative resistance, and warns that even its highest class can mean limited wearing time.

Sewing a waterproof garment punches a line of holes through the barrier. Gore's own guidance is blunt: seam tape seals the tiny holes created by sewing, and although the holes are small they can cause leaks. This is why a jacket with no sealed seams is not a waterproof garment regardless of its fabric, and why taped seams are the single most useful thing you can check by turning a jacket inside out in a shop.

Be careful with the vocabulary, though. Fully taped generally means every seam and critically taped generally means only those the brand judges most exposed, but we found no standard defining either term. What a given brand includes has to be read, not assumed.

Then work through the openings, because that is where water reaches people on long days. A hood needs to adjust at the back of the head as well as around the face, and to turn with you rather than blocking your view. Cuffs need to close over or under gloves. The hem needs to cinch against wind-driven rain. The main zip needs a storm flap or a water-resistant construction. Pockets need to be reachable while wearing a pack's hip belt. None of these is improved by a better membrane, and all of them are visible before you buy.

Ventilation deserves the same attention. Opening a garment moves far more moisture than any membrane, so pit zips or a two-way main zip matter most for high-exertion use — and are dead weight for someone walking a dog in town.

The layer count describes how the barrier is protected on the inside, not how well it works. Pertex offers the same Shield membrane system in 2-layer, 2.5-layer and 3-layer builds, which is proof enough that the number is an assembly choice. Gore states the purpose of its own 2.5-layer construction as packability and low weight, not as a performance tier.

The repellent finish on the outside is not the waterproofing. When the face fabric saturates, Gore states the membrane still keeps the garment waterproof — the jacket simply feels cold and clammy. So a jacket that has stopped beading needs its finish refreshed; it has not necessarily failed.

Platform names also carry less information than buyers assume. Gore's own FAQ says that not all GORE-TEX products are waterproof: only those carrying the black diamond label with its guarantee are described as durably waterproof, while WINDSTOPPER products carry a white diamond and are described as windproof garments suited to a chance of showers. Reading the label variant matters more than recognising the family name.

Weight and toughness pull against each other, and no specification resolves the tension for you. A light face fabric packs small and is pleasant to carry all summer; a heavy one survives a pack and a rock face. Published weights are broadly comparable if you check the size quoted, which is why our records name the size alongside the figure.

Durability is the specification nobody publishes. We have found no fabric maker that quantifies the service life of a laminate, so treat confident durability rankings — including ours — as reasoning rather than measurement. What can be verified is what happens when something does fail: whether the brand operates a repair service, what its warranty actually covers, and whether spare parts and reproofing guidance are published.

On price, the honest framing is that money buys construction, fittings, repairability and often weight savings, rather than a higher number on a hangtag. A jacket that publishes no hydrostatic-head figure can cost several times one that advertises 20,000 mm, and the cheaper garment is not thereby better protected — the two simply disclose different things.

This is our editorial framework rather than a standardised procedure, but it puts the verifiable decisions before the unverifiable ones.

First, describe your worst realistic day: duration, exertion, temperature, wind, and what will rub the garment. Second, set a construction requirement from that — sealed seams always, and hard-wearing fabric only if something will actually abrade it. Third, set a ventilation requirement from exertion, not from the fabric's claimed breathability. Fourth, fit it over the layers you actually own, with the hood up and your arms above your head. Fifth, check the fittings: hood adjustment, cuffs, hem, zip protection, pocket access with a pack on. Sixth, only now compare published figures, and only where the same method is named for both garments. Seventh, check repair and warranty provision. Eighth, buy the garment that satisfies the first five steps most convincingly, not the one with the largest number.

If two candidates survive to the end, the tiebreaker is usually fit and hood quality, because those are the things you cannot fix later.

Our verified records show how uneven disclosure is, which is the strongest argument against shopping by numbers. Arc'teryx publishes no hydrostatic-head figure for the Beta, and neither does Gore for its platform. Patagonia publishes no figure for the Torrentshell 3L either; its H2No Performance Standard page describes 20,000 mm initial and 10,000 mm post-wash thresholds for the standard as a whole, not for that garment.

At the other end, Rab publishes 20,000 mm and 20,000 g/m²/24 h for the Downpour Mountain without naming a test method for either, and Montane publishes 28,000 mm for the women's Phase Nano, also without naming a method. Those are the highest confirmed figures in our database, and we record the missing method alongside each of them.

So a shortlist built purely on published millimetres would exclude several of the most respected shells on the market and rank two unmethodised numbers above them. That is why our jacket records show construction, fittings, weight and warranty provision next to any figure, and label everything a manufacturer does not publish rather than filling the gap.