North or South of the Alps: Which Side Gets the Reliable Snow?
Published 16 Sept 2026
The standard advice is that the north side of the Alps is the snowy side and the Italian side is the dry one, and our own snow-sure ranking repeats a version of it. On 15 winters of modelled history across the 36 resorts we track, that is true as an average and misleading as advice. The nine resorts south of the main Alpine crest have a median season score of 170 against 190 for the 27 on the north side. They also include Breuil-Cervinia, third of 36 at 202, and Val Gardena, last of 36 at 114.
The south side is not one climate, and the line that separates its good resorts from its bad ones runs west to east. The three Aosta Valley resorts score between 198 and 202, level with the Swiss resorts on the other side of the same ridges. The three eastern resorts, Val Gardena, Cortina d’Ampezzo and Madonna di Campiglio, average 129.3 and have the three lowest snowfall-day counts anywhere on the site. Averaging those two groups together is how you get a “dry south” that describes neither of them.
What counts as north and south here
The main Alpine crest is the watershed, so this guide uses the rivers. A resort is on the south side if its valley drains to the Po or straight to the Adriatic, and on the north side if it drains to the Rhine, the Danube or the Rhône. That puts nine resorts south: Breuil-Cervinia, Courmayeur and La Thuile in the Aosta Valley, Sestriere and Sauze d’Oulx in Piedmont, Bormio on the Adda, Madonna di Campiglio on the Sarca, Val Gardena on the Isarco and Cortina on the Piave.
It also puts one Italian resort on the north side. Livigno sits in a valley that drains into the Inn and on to the Danube, so here it counts as north, the same as St Moritz. Both are genuinely borderline, and the section on where the line is drawn below shows what happens if you move them.
The score is the one used across the site’s guides, unchanged: for each of the 14 half-months from November to May, the number of the last 15 winters in which the resort had a 30 cm or deeper base at its mid-mountain elevation by that date, summed to a maximum of 210. Snowfall days are the average count of days with more than 1 cm of new snow, November to May, from the same modelled series.
Is the south side drier?
On average, yes, and by a margin that does not come from altitude.
| Resorts | Median season score | Median mid-mountain | Median snowfall days per season | Median depth, 31 Dec | |
|---|---|---|---|---|---|
| North of the crest | 27 | 190 | 2,200 m | 79.5 | 99 cm |
| South of the crest | 9 | 170 | 2,000 m | 60.2 | 79 cm |
| North, mid-mountain 1,950 m or higher | 21 | 192 | 2,215 m | 81.4 | 111 cm |
The south-side resorts sit lower in the middle, so the first thing to check is whether the 20-point gap is just that. It is not. Every south-side resort here has a mid-mountain of 1,956 m or more, and restricting the north side to the same range widens the gap to 22 points rather than closing it.
Snowfall frequency is where the difference is plainest. The median north-side resort records 19.3 more snowfall days a season than the median south-side one, and the six resorts with the fewest snowfall days on the entire site are all south of the crest. The first north-side resort on that list is Saalbach-Hinterglemm at 63 days, and six of the nine south-side resorts record fewer.
The low scores also fall differently on each side. Five north-side resorts score below 150, and none of them has a mid-mountain higher than 1,760 m: they are low valley resorts, and the altitude guide explains why they cluster. Three south-side resorts score below 150, and none of them has a mid-mountain lower than 2,000 m. On the north side a weak record usually means low ground. On the south side it means something altitude does not fix.
So why is the third best resort on the site Italian?
Because the average hides a split, and the split is the useful part.
| Resort | Drains to | Mid-mountain | Snowfall days per season | Base by 31 Dec | Season score | Rank of 36 |
|---|---|---|---|---|---|---|
| Breuil-Cervinia | Dora Baltea, Po | 2,750 m | 86.2 | 14 of 15 | 202 | 3rd |
| La Thuile | Dora Baltea, Po | 2,200 m | 81.0 | 15 of 15 | 199 | joint 7th |
| Courmayeur | Dora Baltea, Po | 1,956 m | 75.2 | 15 of 15 | 198 | joint 10th |
| Bormio | Adda, Po | 2,000 m | 60.2 | 15 of 15 | 193 | 14th |
| Sauze d’Oulx | Dora Riparia, Po | 2,000 m | 62.2 | 13 of 15 | 170 | 25th |
| Sestriere | Chisone, Po | 2,400 m | 60.0 | 13 of 15 | 169 | 26th |
| Cortina d’Ampezzo | Piave | 2,470 m | 52.4 | 9 of 15 | 139 | joint 30th |
| Madonna di Campiglio | Sarca, Po | 2,000 m | 51.3 | 10 of 15 | 135 | 35th |
| Val Gardena | Isarco, Adige | 2,000 m | 45.4 | 8 of 15 | 114 | 36th |
The spread is 88 points, wider than the whole north side’s 69, from nine resorts in one country on one side of one ridge. Four of the nine score 190 or more, which is a smaller share than on the north side, where it is 15 of 27, but not a small one.
Sort the same nine by position instead of score and a strong pattern appears. Across the nine south-side resorts, longitude correlates with season score at r = -0.74: the further east, the weaker the record. Mid-mountain altitude, over the same nine, correlates at r = 0.21, which is barely a relationship at all. On the south side, knowing where a resort is tells you far more than knowing how high it is.
At the two ends of that line the numbers barely overlap. The Aosta Valley trio has a median of 199 and a spread of 4 points, and averages 80.8 snowfall days a season. The eastern trio has a median of 135, averages 49.7 snowfall days, and cleared a 30 cm base by 31 December in a median 9 of 15 winters against 15 of 15 in Aosta. On that date the median depth is 132 cm in the Aosta trio and 49 cm in the eastern one. Cervinia’s 86.2 snowfall days a season is the third highest count on the entire site, behind only the Stubai Glacier and Andermatt on the north side.
The two Piedmont resorts are the exception to that pattern, the furthest west of all nine and still only 25th and 26th: the altitude guide already covers Sestriere as a case of height compensating for a thin supply, and the Aosta result does not extend south-west into Piedmont. Bormio is the stranger result, 14th of 36 on only 60.2 snowfall days, which is the same fewer-but-better-banked pattern that guide describes for it at 2,000 m.
What happens when you cross the ridge?
The cleanest test available is to pair each south-side resort with its nearest north-side neighbour. Where that neighbour is close, it is as near to a like-for-like comparison as the dataset gets.
| South side | North side | Distance | Mid-mountain, south vs north | Score, south vs north |
|---|---|---|---|---|
| Breuil-Cervinia | Zermatt | 13 km | 2,750 m vs 2,582 m | 202 vs 205 |
| Bormio | Livigno | 19 km | 2,000 m vs 2,300 m | 193 vs 186 |
| Courmayeur | Verbier | 40 km | 1,956 m vs 2,200 m | 198 vs 200 |
| La Thuile | Verbier | 47 km | 2,200 m vs 2,200 m | 199 vs 200 |
| Val Gardena | Obergurgl-Hochgurgl | 63 km | 2,000 m vs 2,150 m | 114 vs 191 |
| Madonna di Campiglio | Livigno | 63 km | 2,000 m vs 2,300 m | 135 vs 186 |
| Cortina d’Ampezzo | Mayrhofen | 73 km | 2,470 m vs 1,565 m | 139 vs 155 |
Distances are straight lines between resort coordinates, and pairs skip the Hintertux glacier so that neither resort is a glacier: Val Gardena pairs with Obergurgl, and Cortina with Mayrhofen, 0.1 km further than Hintertux. Sestriere and Sauze d’Oulx are left out because their nearest north-side resort on this site is over 120 km away, which is not a pair.
In the west, crossing the crest costs nothing. Cervinia and Zermatt share the same lift-linked ski area over the Theodul and finish three points apart. La Thuile and Verbier sit at an identical 2,200 m and finish one point apart. And in the single closest pair on the list, further east in Lombardy, the Italian side wins: Bormio scores seven points higher than Livigno from 300 m lower down, 19 km away over one pass.
In the east, the same crossing is worth 51 points between Madonna di Campiglio and Livigno, and 77 between Val Gardena and Obergurgl, where the Austrian resort is only 150 m higher. Cortina loses by 16 to Mayrhofen even though its mid-mountain sits about 900 m higher.
The usual explanation is shelter. The western Italian resorts sit directly under the highest massifs in the Alps, Mont Blanc and the Matterhorn, where moist air arriving from more than one direction is forced up and unloads, while the eastern resorts sit deeper inside the mountain mass, with more ranges between them and the weather. That is a meteorological reading, not something this dataset measures, and we would rather say so than present it as a result.
Does it matter where the line is drawn?
For the average, a great deal. For the finding, not at all.
Livigno and St Moritz are on the north side here because of where their rivers go, but both are high Italian-border valleys often grouped with the south. Move them across and the south-side median rises from 170 to 186, because both score well. Take the other borderline group, the dry inner-Alpine valleys of the Valais and the Engadin plus Livigno, out of the north side instead and the north-side median falls from 190 to 188.
A 16-point swing from reclassifying two resorts is a clear sign that “north side” and “south side” are weak categories for predicting snow. The west-east split inside the south does not depend on any of those choices: none of the Aosta or eastern resorts is borderline, and the 13 km Cervinia and Zermatt pair and the 63 km Val Gardena and Obergurgl pair do not change with the definition.
Does the south side run a later season?
The March guide says the southern side runs later, and for the eastern resorts the data agrees. For the south side as a whole it does not.
The share of a season’s snowfall days that falls in November and December is 28.6% for the median north-side resort and 28.8% for the median south-side one, which is no difference at all. The eastern trio is the outlier at 26%, and across all 36 resorts only four record more snowfall days in March than in January: Val Gardena, Cortina, Madonna di Campiglio and, on the north side, Obergurgl-Hochgurgl. The Aosta resorts follow the northern calendar.
The eastern trio’s weakness is also concentrated at the ends of the season rather than spread through it. In the first half of February its median is 15 of 15 winters with a 30 cm base, the same median as everywhere else. By 30 April it is 7 of 15, against 15 of 15 for the Aosta trio. The eastern resorts are a February and early-March booking on natural snow, and a gamble either side of that.
Is Italy less snowy than Austria, then?
Not on this data, and it is worth saying because it is the form most people ask the question in. Italy’s ten resorts have a median season score of 178 and Austria’s sixteen 177.5, with median mid-mountains of 2,100 m and 2,000 m. Italy has both a much stronger top and a much weaker bottom, and the country average cancels them out. Switzerland’s figures are higher, but as the altitude guide sets out, we built the biggest and highest Swiss resorts first, so a country comparison including Switzerland is a sampling question before it is a snow one.
What this data cannot tell you
It is modelled, not measured. The history comes from the Open-Meteo reanalysis archive at one mid-mountain point per resort. Seven Austrian resorts have a GeoSphere Austria station nearby to sanity-check against; the Italian and Swiss ones have no equivalent check on this site, which matters more in a guide whose whole subject is the Italian side. Zermatt’s history is sampled 6 km down its own valley near Täsch, because the cell over the resort carries permanent glacier ice, so the Zermatt side of the closest pair is a proxy point.
Snowmaking is not modelled. The scores describe natural snow, not the pistes you would ski, and the resorts this flatters least are the ones at the bottom of the table.
Snowfall days are frequency, not amount. A day counts if more than 1 cm fell. Bormio’s high score on few snowfall days is the plainest reminder that one heavy storm and six light ones count the same way.
Nine resorts is a small side. Every south-side median above rests on nine numbers, and two of the most useful groups here are three resorts each. The pattern is strong, though not uniform (the western Piedmont pair sits mid-table), and a tenth resort in the wrong place would move the medians, and the Italian south side has large ski areas this site does not yet cover.
Fifteen winters is a deliberately short window, recent enough to describe the current climate rather than an old average. Peer-reviewed station analysis (Matiu et al. 2021) found the steepest snow-depth declines in the Alps at lower elevations and in the shoulder seasons, which is where the eastern resorts are already weakest.
How to use this
- Do not choose a side, choose a region. The south side contains both the third best record on this site and the worst. “Italian Alps” and “north of the Alps” are not useful filters on their own.
- In the west, treat the border as invisible. Breuil-Cervinia, La Thuile and Courmayeur score within four points of each other and within three of their Swiss neighbours. Book them on terrain, price and access.
- In the east, book the middle of the season. The eastern resorts match the site-wide median in the first half of February and fall away sharply by the end of April. For December or late April, the same money on the north side of the crest buys a much safer bet.
- Check Bormio before writing off Lombardy. It out-scores Livigno from 300 m lower, on fewer snowfall days, and cleared a 30 cm base by 31 December in 15 of 15 winters.
- Then read the half-month on the resort page, because a season score is a sum and your holiday is one or two of its parts.
The short version: the south side of the Alps is drier on average, but the average is made of two different places. The Aosta Valley holds its snow as well as Switzerland does, the eastern resorts hold it worse than anywhere else we track, and the ridge between north and south matters far less than which end of it you are standing at.