Waterproof Ratings Explained

Every waterproof rating you see on a hangtag — 5,000 mm, 10,000 mm, 20,000 mm, occasionally 30,000 mm — is a hydrostatic head result. A fabric sample is clamped, water pressure is applied to one face and raised at a controlled rate, and the pressure at which water penetrates is recorded. Because the pressure is generated by a column of water, the natural unit is the height of that column.

So the unit is a pressure, not a depth of immersion and not a duration. 10,000 mm of water column is the pressure at the base of a ten-metre-deep column of water; 20,000 mm is twenty metres. Under ISO 811:2018, the international method most often behind these figures, the recorded pressure is the point at which water appears in the third separate place on the specimen, not the first droplet — the fabric is already failing in three spots when the number is taken.

The scale is worth pausing on, because it explains why the numbers sound more decisive than they are. Rain arrives as droplets carrying almost no pressure. Nothing in weather produces a ten-metre water column against your shoulder. The pressure that genuinely occurs in use is mechanical and local: a loaded pack strap pressing wet fabric into your shoulder, a hip belt, a knee in wet grass, sitting on a soaked rock.

ISO 811 exposes an area of 100 cm² of clamped fabric, tests at least five specimens taken from different places in the material, and reports both individual results and their mean. Its scope is fabric, and the standard limits its own relevance to articles subjected to water pressure for short or moderate periods.

That leaves out almost everything a jacket is made of. Every seam is a line of needle holes, sealed — or not — with tape bonded over the stitching. The main zip, hood opening, cuffs, hem and pocket entries are all deliberate holes in the barrier. None of them appear in a fabric rating. Nor does the face fabric's water-repellent finish, which is a separate surface treatment: when it wears off, the outer textile soaks up water and the jacket feels cold and clammy, and Gore states plainly that in that state the membrane still keeps the garment waterproof.

Manufacturer vocabulary around seams deserves the same scepticism as the numbers. "Fully taped" and "critically taped" are brand terms; we found no standard defining either, so they describe intent rather than a verified level of coverage.

Consumer millimetre claims are unregulated marketing disclosures. There is a regulated benchmark for rain protection, though, and it is the one place where a threshold can be quoted as a requirement rather than an opinion: the European standard for protective clothing against rain, used for professional wet-weather workwear.

That standard changed identity recently, and several outdoor sites have not caught up. ISO published ISO 24232:2024 in October 2024, and the European adoption EN ISO 24232:2025 states on its cover page that it supersedes EN 343:2019. The CEN register records EN 343:2019 as withdrawn with effect from 22 October 2025. The class thresholds themselves did not move: reading table 2 of ISO 24232:2024 directly, class 1 requires at least 8,000 Pa for material and seams before pre-treatment, class 2 the same level after pre-treatment, class 3 at least 13,000 Pa, and class 4 at least 20,000 Pa with seams also tested after cleaning.

Two details in that clause matter more than the numbers. Where specimens from one garment fall into different classes, the lowest result classifies the whole garment. And the standard carries an explicit warning that all classes — including the highest — can result in a restricted wearing time. Even the regulated top class is not a promise of indefinite dryness.

The standard also reaches the finished garment in a way consumer figures never do: it sets requirements for seams as well as materials, and includes an optional clause for the waterproofness of a ready-made garment, with a separate ISO method for high-energy droplets falling from above. Optional is the key word — when it has not been done, nothing about garment-level performance has been established.

There is no single hydrostatic head method. ISO 811 is one; AATCC TM127 is the North American method, performed without a restraint, with a restrained variant published separately as TM208; ASTM D751 covers coated fabrics and permits two different pieces of apparatus within one standard, a Mullen type tester and a rising water column tester; JIS L 1092 splits the test into a low-pressure and a high-pressure method, the latter for specimens tested at 10 kPa or more.

Even within ISO 811, the permitted conditions change the answer. Two rates of pressure increase are allowed, and the standard states results at different rates may not be the same. Two water temperatures are permitted, and the standard notes the higher one produces lower values. Which face of the fabric was tested, and whether the specimen was new or pre-treated, change it again.

So when two brands both print 20,000 mm and neither names a method, they have told you they each ran a test they were satisfied with. They have not told you the results are equivalent. We do not convert between methods anywhere on this site, because no standard we can read authorises it — and treating a figure from one method as directly better than a lower figure from another is exactly the error this page exists to prevent.

Among our verified records, Rab publishes 20,000 mm of hydrostatic head for the men's Downpour Mountain, and Montane publishes 28,000 mm for the women's Phase Nano. Neither names the test standard, the pressure rate or the water temperature.

Patagonia shows why the figure is a moment in time rather than a fixed property. Its H2No Performance Standard tests fabrics at 20,000 mm, subjects them to a wet-flex and abrasion routine the brand says simulates up to 30 years of drenching use, then requires the fabric to still hold 10,000 mm. The same fabric honestly carries both numbers, before and after — which is a far more informative disclosure than a single figure.

And for a large part of the market the number does not exist. GORE-TEX publishes no hydrostatic head figure for its laminates, certifying finished garments against its own wet-weather requirement instead; the Arc'teryx Beta page carries no waterproof or breathability figure; Helly Hansen publishes no numeric rating for the three-layer Odin 9 Worlds 3.0. Those absences are disclosure choices, not performance verdicts, and we do not fill them in.

The best-known table in outdoor retail maps millimetres onto weather: 5,000 mm for light rain, 10,000 mm for moderate rain, 20,000 mm for heavy rain, 30,000 mm for extreme conditions. We looked for its origin in the standards that define the measurement and in manufacturer documentation, and found nothing that establishes those thresholds. The measurement is a pressure; rain intensity is a rate of fall. They are not the same quantity, and no method we can read converts one to the other.

The one defensible reference point is the regulated class structure above: at least 816 mm of water column for the lowest waterproof class and roughly 2,040 mm for the highest, in both cases after the pre-treatments the standard requires, with seams included. Every common consumer figure sits well above that. Where you see a rain chart on another site, treat it as retail shorthand rather than evidence.

This is our interpretation rather than a measurement, and we would apply it in this order. First, seam sealing: tape over every stitch line, which you can see by turning the jacket inside out. Second, the openings — hood adjustment and brim, cuff closure, hem drawcord, and whether the main zip has a storm flap or a water-resistant construction. Third, the face fabric and how easily you can wash and re-treat it, because repellency is the property that degrades first and is most often mistaken for leaking. Fourth, ventilation, because on any hard effort your own moisture is the more likely cause of a wet inside.

The millimetre figure comes after those, and it is most useful in one narrow situation: sustained mechanical pressure on wet fabric, and as headroom for a fabric to age into. If two jackets are otherwise equal and one is rated higher, that is a reasonable tie-breaker. It is a poor first filter.