Hostel supply census
Most cities have no hostels. We counted hostel and backpacker accommodation within about five miles of the centre of 2,876 cities with an event venue, on 18 September 2026. 2,414 of them (83.9%) had none at all, and a further 281 had fewer than five. Only 116 cities hold enough for a real choice, between them 3,231 hostels.
Every travel site will show you a hostel list wherever you search. That hides the actual distribution, which is extremely lopsided: a few dozen cities have scores of them and the large majority have zero. This is the whole survey, zeroes included, free to download and free to reuse.
What the survey found
| Measure | Cities | Share of those surveyed |
|---|---|---|
| Surveyed | 2,876 | 100% |
| No hostels at all | 2,414 | 83.9% |
| One to four hostels | 281 | 9.8% |
| Five or more (own page) | 116 | 4.0% |
| Hostels counted in total | 3,231 | across the qualifying cities |
Counts are properties listed by our accommodation partner within about five miles of each city centre, measured in a straight line, as read on 18 September 2026. A count of zero is a measurement, not missing data. Nothing on this page is a price.
Cities with the most hostels
The 40 densest of the 116 cities that cleared the threshold. Guest scores are our accommodation partner’s, out of 10. They are not StubOrder reviews.
| City | Hostels | Share of stays | Hostel score | Hotel score |
|---|---|---|---|---|
| Madrid, Madrid | 156 | withheld | 7.3 | 8.0 |
| Barcelona, Catalonia | 151 | 3.9% | 7.6 | 8.1 |
| Lisbon, Lisbon | 130 | withheld | 7.1 | 8.4 |
| London, Greater London | 107 | withheld | 7.0 | 8.6 |
| Maracana, Rio de Janeiro | 88 | 2.2% | 7.9 | 8.2 |
| São Paulo, Sao Paulo | 83 | 2.0% | 8.0 | 8.3 |
| Kuala Lumpur, Kuala Lumpur | 79 | 1.6% | 7.2 | 7.7 |
| Prague, Praha | 78 | 2.6% | 7.7 | 8.7 |
| Baku, Azerbaijan | 77 | 3.8% | 6.3 | 7.7 |
| Krakow, Poland | 73 | 2.9% | 7.9 | 8.3 |
| Warsaw, Poland | 72 | 2.4% | 7.0 | 7.6 |
| Porto, Oporto | 71 | withheld | 8.2 | 8.2 |
| Budapest, Budapest | 70 | withheld | 7.5 | 8.9 |
| Bogota, Colombia | 68 | 2.9% | 7.0 | 7.7 |
| Mexico City, Ciudad de Mexico | 68 | 2.3% | 7.5 | 8.0 |
| Quito, Ecuador | 66 | 8.2% | 7.2 | 8.3 |
| Belgrade, Belgrade | 65 | 1.7% | 7.9 | 9.1 |
| Santiago, Santiago RM | 58 | 2.4% | 7.4 | 8.2 |
| Singapore, Singapore | 58 | 9.7% | 7.2 | 7.7 |
| Vienna, Vienna | 58 | 1.3% | 7.3 | 8.1 |
| Berlin, Berlin | 54 | 3.3% | 7.5 | 8.2 |
| Rome, Rome | 47 | withheld | 7.6 | 9.0 |
| Amsterdam, North Holland | 44 | 3.7% | 7.3 | 8.4 |
| Gdansk, Poland | 44 | 1.2% | 8.0 | 8.1 |
| Milano, Milan | 44 | withheld | 7.2 | 9.0 |
| Los Angeles, California | 42 | 5.7% | 6.8 | 7.3 |
| Sevilla, Andalusia | 41 | 1.0% | 7.7 | 8.6 |
| Malaga, Andalusia | 40 | 0.8% | 7.7 | 8.5 |
| Tirana, Albania | 40 | 1.3% | 8.0 | 9.0 |
| Paris, Ville de Paris | 36 | withheld | 7.6 | 8.0 |
| Zagreb, Croatia | 36 | 1.5% | 7.9 | 9.1 |
| San Jose, Costa Rica | 35 | 7.5% | 8.4 | 8.3 |
| Athens, Attica | 34 | withheld | 7.8 | 9.2 |
| Hamburg, Hamburg | 32 | 4.4% | 7.1 | 8.1 |
| Miami, Florida | 32 | 1.1% | 5.4 | 7.7 |
| Valencia, Valencia | 32 | 1.4% | 7.5 | 8.4 |
| Zapopan, Jalisco | 31 | 4.1% | 7.2 | 8.0 |
| Auckland, Auckland | 29 | 2.1% | 7.6 | 7.9 |
| Dublin, Dublin | 29 | 3.2% | 6.9 | 7.8 |
| Edinburgh, Edinburgh | 29 | 1.4% | 7.4 | 8.2 |
“Withheld” means that city’s total property count hit the survey’s fetch limit, so its denominator is known to be too small and no honest share can be computed from it. 22 cities are withheld.
Where hostels are the biggest share of what exists
A different question, and a different ranking. A city can hold a hundred hostels and still be overwhelmingly a hotel city; this column is what separates a genuine backpacker destination from a large one.
- Singapore, Singapore: 9.7% of the 598 places to stay nearby, 58 hostels
- Quito, Ecuador: 8.2% of the 805 places to stay nearby, 66 hostels
- San Jose, Costa Rica: 7.5% of the 464 places to stay nearby, 35 hostels
- Mannheim, Baden-Württemberg: 7.0% of the 314 places to stay nearby, 22 hostels
- Los Angeles, California: 5.7% of the 733 places to stay nearby, 42 hostels
- Gliwice, Poland: 5.4% of the 111 places to stay nearby, 6 hostels
- Stockholm, Stockholm: 5.3% of the 525 places to stay nearby, 28 hostels
- Goteborg, Vastra Gotaland: 5.0% of the 261 places to stay nearby, 13 hostels
- Hamburg, Hamburg: 4.4% of the 720 places to stay nearby, 32 hostels
- Zapopan, Jalisco: 4.1% of the 764 places to stay nearby, 31 hostels
- Wellington, Wellington: 4.1% of the 246 places to stay nearby, 10 hostels
- Barcelona, Catalonia: 3.9% of the 3,911 places to stay nearby, 151 hostels
- Perth, Western Australia: 3.9% of the 510 places to stay nearby, 20 hostels
- Baku, Azerbaijan: 3.8% of the 2,001 places to stay nearby, 77 hostels
- San Francisco, California: 3.8% of the 391 places to stay nearby, 15 hostels
How it was measured
- Scope. Every city with an event venue on StubOrder whose coordinates we hold, 2,876 of them. Not a sample.
- Radius. About five miles (8km) from the city centre, straight line. Not a driving distance and not a city-limits boundary.
- What counts. Properties our accommodation partner classifies as a hostel or as backpacker accommodation. Their classification, so a cheap hotel does not count as a hostel.
- The denominator. All accommodation of any type the same search returned in the same radius. Where that count saturated the fetch limit, the share is withheld entirely.
- Guest scores. The supplier’s, out of 10. We do not average them into a rating of our own and publish no aggregateRating for them.
- Merged cities. Where two adjacent cities return the same properties (a city and the suburb next to it), the smaller is marked as merged so the same hostels are not read twice. The CSV carries that column.
- What is missing. This is one reading on one date. It says nothing about prices, bed counts, opening or closing, or quality beyond the supplier’s own score.
Using this data
Published under CC BY 4.0: download it, chart it, quote it, build on it. Credit StubOrder and the survey date (18 September 2026) so a reader can find the version you used. If a figure looks wrong we would rather hear about it. The contact form reaches a person, and a correction changes how the figure is derived everywhere it appears.
Related
Frequently asked questions
Can I reuse this data?
Yes. It is published under CC BY 4.0: download the CSV, use it however you like, and credit StubOrder and the survey date. If you are writing something and want the method explained, the methodology page covers it and the contact form reaches a person.
What counts as a hostel here?
A property our accommodation partner classifies as a hostel or as backpacker accommodation. It is their classification, not ours, and it means a budget hotel does not creep into the count just because it is cheap.
How many cities have no hostels at all?
2,414 of the 2,876 cities we surveyed - 83.9% - had no hostel or backpacker accommodation listed within about five miles of the centre when we looked on 18 September 2026. A further 281 had between one and four, too few for the choice to be real.
Are these hostel prices?
No. There are no prices in this dataset at all. A nightly rate only exists for specific dates, so there is no standing price for a hostel or for a city. Everything here measures supply and guest scores.
Why do some cities have no share percentage?
Because their total property count hit the survey's own fetch limit, so the denominator is known to be too small and any share computed from it would be too high. 22 cities are withheld on that basis. We would rather print nothing than a number that is wrong in a predictable direction.
