For each of 30 French resorts we take daily snow depth for the last 10 winters (2016–17 to 2025–26) from Open-Meteo, at two-thirds of the way up the resort. Every day is ranked against the rest of that resort's season, so a week is judged against the resort itself, not against a fixed number of centimetres.
Snow odds 8/10 means: in 8 of the last 10 winters, these dates were in the snowier half of that resort's season. It is history, not a forecast.
When we compare resorts (all of France, or "Best resorts for these dates"), a week's snow counts half against the resort's own season and half against the other resorts on the same dates, so a deep resort's ordinary week is not outranked by a thin resort's best one.
Crowds come from the 2026–27 school holidays of France (zones A, B and C), England, Belgium and the Netherlands, plus weekends. Each country counts once, however many holiday zones it has. Quiet, busy and very busy are thresholds on that mix.
The score out of 100 mixes snow and quiet slopes with the weights you choose (60/40 by default).
Before launch we asked whether history picks a good week: for each past winter we picked the best week from the other nine and checked it against what really happened. The pick landed in that winter's top quarter of weeks 58% of the time and beat a simple "go in early February" rule in 7 of 10 winters. Good, not magic. That is why we show odds rather than promises.
Snow: Open-Meteo historical weather (CC BY 4.0). Resorts: OpenSkiMap, © OpenStreetMap contributors (ODbL). School holidays: OpenHolidays; England from council term dates. Weather-model snow is smoothed over about 10 km, so treat centimetres as a guide; the ranking of weeks is what the odds use.
No cookies, no accounts, no ads. We count searches anonymously (no IP, no identifier) to see whether the site is useful.