Hakuba's Worst Winter in Twelve Years Brought 301 cm. Niseko's Brought 166. Book on the Floor, Not the Swing.

Corrected on 13 September 2026. An earlier version of this article compared resorts by the gap between their best and worst of three winters, and concluded that Kiroro (9%) was far steadier than Hakuba (62%), that Hokkaido and northern Tohoku were Japan’s reliable snow, and that Appi Kogen and Hakkoda offered “Hokkaido-grade consistency”. Measured over twelve winters, none of that holds: Kiroro’s gap is 70% and Hakuba’s 74%, and Iwate went from Japan’s steadiest prefecture to one of its least steady. We have rewritten the article around a better question. What went wrong is at the end.

Booking a ski trip to Japan means committing in August to conditions in February. Flights, accommodation and leave are locked in long before anyone knows what kind of winter it will be.

So the useful question is not how much snow a place gets on average. It is: if this turns out to be a bad year, how bad does it get here?

We now have twelve winters of snowfall for all 377 Japanese ski areas that average 100 cm or more. That is enough to answer the question directly, by looking at each resort’s worst winter.

Japan’s bad winters happen everywhere at once

Of the 377 ski areas, 240 had their worst winter of the twelve in 2015/16. Another 68 had it in 2019/20. Between them, those two winters account for more than four in five.

That changes the strategy. If bad winters were local, you could hedge by choosing a “steady region”. They are not local. In a bad year, every region you might have picked is having it too. What you are really choosing is how much snow is left when that happens.

The famous resorts, by their worst winter

	Worst of 12 winters	Average	Best of 12	Worst as % of average
Nozawa Onsen	301 cm	450 cm	635 cm	67%
Hakuba Happo-One	301 cm	414 cm	609 cm	73%
Kiroro	209 cm	331 cm	442 cm	63%
Shiga Kogen	197 cm	255 cm	337 cm	77%
Furano	181 cm	220 cm	266 cm	82%
Niseko Grand Hirafu	166 cm	289 cm	384 cm	57%
Rusutsu	155 cm	212 cm	291 cm	73%
Zao Onsen	129 cm	258 cm	357 cm	50%

Hakuba’s and Nozawa’s worst winters of the last twelve delivered more snow than Niseko’s average winter. That is the whole argument in one line.

Furano keeps the largest share of its normal snow in a bad year — 82%, which makes it the steadiest of the eight. But 82% of 220 cm is 181 cm. Steadiness and depth are different columns, and for a trip you cannot move, the one that matters is the floor.

Where the floor is deepest

The ski areas whose worst winter of the twelve was still the deepest:

Worst of 12 winters	Average
388 cm	690 cm	Charmant Hiuchi	Niigata
385 cm	553 cm	Togakushi	Nagano
385 cm	635 cm	Hakuba Cortina	Nagano
381 cm	765 cm	Okutadami Maruyama	Niigata
380 cm	668 cm	Hakuba 47 · Tsugaike	Nagano
370 cm	655 cm	Gassan	Yamagata
353 cm	448 cm	Hakuba Goryu	Nagano
328 cm	712 cm	Yuzawa · Iwappara	Niigata

Every one of these took more snow in its worst winter than most Japanese ski areas take in an average one. They are concentrated in the Niigata–Nagano snowbelt and the Hakuba valley, with Gassan as the northern exception.

Togakushi is worth a separate line: 385 cm in its worst winter against a 553 cm average, so it keeps 70% of its normal snow in a bad year and starts deep. It also sits inside Nagano city.

By prefecture

Prefecture	Ski areas	Worst winter, typical	Average, typical	Worst as % of average
Niigata	47	306 cm	567 cm	53%
Nagano	60	214 cm	327 cm	64%
Gunma	23	183 cm	252 cm	63%
Akita	17	182 cm	280 cm	58%
Iwate	11	167 cm	332 cm	51%
Aomori	9	154 cm	234 cm	66%
Yamagata	17	151 cm	312 cm	51%
Fukushima	25	151 cm	325 cm	51%
Hokkaido	86	149 cm	209 cm	71%
Gifu	16	136 cm	220 cm	64%
Miyagi	11	111 cm	190 cm	67%
Hiroshima	8	57 cm	116 cm	50%
Hyogo	10	51 cm	205 cm	27%

Hokkaido loses the smallest share of its snow in a bad year, and still ends up with less than half of Niigata’s. Its typical ski area keeps 71% of normal; Niigata’s keeps 53%. But 71% of 209 cm is 149 cm, and 53% of 567 cm is 306 cm.

This is the same pattern we found counting March snow: the percentage and the centimetres point in opposite directions, and you ski on the centimetres.

Hyogo is the warning at the bottom. Its typical ski area gets a respectable 205 cm on average and 51 cm in its worst winter — a quarter of normal. That is what a bad year looks like where the snow depends on being just cold enough.

What this changes about booking

If your dates are fixed, book deep, not steady. A resort that loses half of a very large winter still leaves you more than one that loses a fifth of a small one. The Hakuba valley, Nozawa, the Yuzawa cluster and Okutadami all had worst winters above 300 cm; Myoko’s deepest ski areas bottomed out at 273 cm.

Hokkaido is the steadier bet in percentage terms, not in snow. If what you want is dry, cold snow, which Hokkaido does better than anywhere in Japan — go for that reason. Do not go because you expect it to hold up in a bad year better than Nagano. In the bad years it held a larger share of less.

Northern Tohoku is not the insurance policy it looked like. An earlier version of this article recommended Appi Kogen, Hakkoda and Akita Hachimantai for consistency. Over twelve winters their gaps run 85–89%, among the widest in Japan. They remain deep places in normal years; they are not steady ones.

If you can move your dates late, none of this matters much. Book when the season has shown what it is doing, and go where it has snowed.

Why this article changed {#why-this-article-changed}

The first version used the best and worst of the three most recent winters, because three winters was what our seasonal snowfall figures used at the time.

A season average survives a short record. A range does not. Three numbers that happen to land close together — as Kiroro’s did, 325 to 357 cm — look like a steady resort. Kiroro’s range across twelve winters is 209 to 442 cm.

It is worse than noise at single resorts, because every ski area shares the same three winters. If those three were all ordinary years, the whole country looks steady at once. Our three-winter figure for Japan as a whole was a 35% gap; the twelve-winter figure is 78%. Neither 2015/16 nor 2019/20 was among the three winters we had used, and those are the two winters that define Japan’s floor.

We have made the same correction on our Andes articles, where the problem was larger still. An earlier version of this article also compared Japan with other countries on the three-winter basis; that table has been removed until we have re-measured the other countries over the same twelve winters.

How this was measured

Daily snowfall from the ERA5 reanalysis, summed over each of the twelve winters from 2014/15 to 2025/26 (December to March), at each ski area’s coordinates, for the 377 Japanese ski areas averaging 100 cm or more. The average is the mean of the twelve; the worst and best are the lowest and highest of the twelve. Prefecture figures are medians across ski areas, for prefectures with eight or more of them.

Neighbouring resorts share a figure. Hakuba 47, Iwatake and Tsugaike read identically because the reanalysis grid does not separate them. Treat a valley as one entry, not three.

These figures run low. ERA5 works on a 10–30 km grid, which flattens the mountains, so mountain snowfall comes out below what actually falls. Checked against Japan Meteorological Agency stations it runs low everywhere but not evenly: on the Sea of Japan side of Honshu it comes close — the station 2.3 km from Hakuba Happo-One records 447 cm against the model’s 453 cm for the same three winters, and in Hokkaido it returns about half, because the model converts precipitation to snow at a fixed ratio and Hokkaido’s colder, drier snow piles up deeper for the same water.

That bias runs in Hokkaido’s favour here, not against it. If Hokkaido’s true snowfall is higher than the model says, its true floor is higher too, and the gap to Niigata is smaller than these tables show. It would take roughly doubling Hokkaido’s figures to close it, which is about the size of the station discrepancy, so read the Hokkaido–Niigata comparison as close, and the comparison between Hokkaido and Honshu’s deepest valleys as the part that holds.