Short-Term Rental Statistics for 2026: What the Numbers Actually Tell Hosts

The report counts 8,347,967 public Airbnb listings across 218 active countries and territories. The top five account for about 47.6% and the top ten about 62.6%; the United States alone holds 18.22%, and the US and France together hold 30.0%. A top-ten cutoff summarizes this snapshot; it is not a universal definition of which countries are the world’s “major” short-term-rental markets.

The UK is fifth by Airbnb listing count (379,001; 4.54%), only 12,830 listings ahead of Spain. A different column changes the comparison: Inside Airbnb estimates 23.1m UK listing-nights versus 22.1m in Italy, using a review-based model. That is a modelled proxy, not an observed booking rank.

A different view comes from AirDNA’s public cross-platform ranking: London is first among its 12 largest published markets, with 64,432 active listings and 63% trailing-12-month occupancy. The directory was updated on 29 September 2026 and includes 36,218 markets in 69 countries. That commercial, source-defined market ranking does not change the UK’s fifth place in the separate Airbnb-only country snapshot.

At the regional level, Europe (43.06%) and the Americas (37.71%) have the largest listing shares—80.77% combined. The country-level leaders show how the regional totals are distributed:

Source: Inside Airbnb / Housing Justice Data Lab report and methods. These are the largest countries by absolute Airbnb listing count in the report, not a claim that regional conditions are interchangeable. The UK is not Europe’s largest country market—France is—but the global ranking places the UK fifth overall.

The per-resident view changes the picture again: the snapshot counts 54.9 UK listings per 10,000 residents, versus 44.3 in the US and 45.0 in Canada; France is 148.1, Italy 89.0, Spain 76.4, and Australia 62.3. Absolute supply and listings per resident answer different questions. France is far larger in raw count, while several smaller markets have higher listing density.

Read the measures separately: the 8.35m count covers Airbnb listings publicly visible during October–December 2025—not Vrbo, Booking.com, direct bookings, or every professionally managed rental. The annual listing-night estimates are inferred from reviews using a 50% review-to-booking assumption and a 70% occupancy cap. The report is housing-advocacy research. Its listing counts support a dated Airbnb supply comparison; its modelled nights are only a directional, within-source estimate.

Three units, one booked stay: guest-nights count each traveller for each night; occupied property nights count nights an accommodation was occupied; a stay counts one booked accommodation stay. Four guests sharing one accommodation for three nights create one stay, three occupied property nights, and 12 guest-nights. The UK figures in the visual are alternative counts of the same activity, not numbers to add together.

What the platform total says—and does not say: Eurostat reports 951.6m guest-nights for 2025. Its 1 April news release labels the aggregate “EU”; the detailed annual article says “EU and EFTA” and elsewhere refers to the same total as EU. The source scope is unresolved, so do not label this a confirmed EU-27 total. In the detailed country table, France leads the EU at 213m, followed by Spain (189m), Italy (139m), Germany (68m) and Greece (52m). The UK ONS separately records 100.9m guest-nights; it is a separate national series, not a row in a harmonised Eurostat country ranking.

UK activity grew and was mostly domestic: ONS records 11.5% annual growth in UK guest-nights and says 67.2% were from domestic visitors in 2025. The national total is strong evidence that the UK is a substantial booked-activity market; the mix matters too if a property plan assumes international demand.

Keep the provider change in view: Eurostat says 2025 is the first year after Tripadvisor exited the panel. The published 11.4% year-on-year rise therefore compares a changed provider mix; it should not be read as a pure like-for-like growth rate. ONS’s separate UK series reports guest-nights up 11.5%.

For a property decision: a host can sell permitted, available property nights—not regional guest-nights. Neither national series tells you the nights or income available to one address. Check local monthly bookings, property type, legal limits, costs, and the capacity to serve guests.

The two ranked lists reveal different things. Inside Airbnb’s country snapshot puts the US first, France second and the UK fifth by public Airbnb listings; its top ten countries account for 62.6% of that dataset. AirDNA’s current top 12 ranks local markets across three booking platforms. London leads its published list, followed by Orlando, Paris, São Paulo and Buenos Aires.

Market activity is not the same as a country’s total supply. Canada and Australia are in Inside Airbnb’s top ten countries but have no market in AirDNA’s published global top 12. Japan and Argentina appear in AirDNA’s top 12 but not Inside Airbnb’s top ten. The rankings do not conflict: one counts Airbnb listings by country, the other counts provider-defined local markets across three platforms.

AirDNA reports both occupancy and average daily rate because listing volume alone does not describe how a market performs. Even those local figures are not a host’s expected return: occupancy is based on nights available to book, and average daily rate covers booked nights before operating costs.

The coverage limits matter too. Inside Airbnb identifies India and South Africa as the largest country-level listing markets in Asia and Africa in its dataset. AirDNA’s free directory lists only 20 Indian markets across one state; that page cannot establish national Indian occupancy. A country absent from a provider’s global top 12—or a country represented by a narrow page—should not be read as having no meaningful short-term-rental activity.

Seasonality: compare like with like

A summer-heavy property has fewer months to cover annual costs. In the same Eurostat series, 57% of Croatia’s platform guest-nights fell in July and August, against 33.1% for the EU overall and 22% for Luxembourg. These shares show when stays occurred; they are not occupancy rates or nightly prices.

A national market still concentrates locally: the 20 most-visited EU regions accounted for 47.1% of EU platform guest-nights. In the UK, 24.4% of guest-nights occurred in just 9 of 361 local authorities; London had the largest total. ONS also reports that 67.2% of UK guest-nights were domestic, compared with 89.5% in Wales and 55.8% in Scotland; about 1% of UK records had unclassified origin. The national total therefore contains materially different local and visitor markets.

The latest UK performance pulse: VisitBritain’s August 2026 update for June says nights reserved and occupancy were stable year on year, while average daily rates and average revenue per property reached new June highs. It withdrew the report’s 2026 supply figures while investigating a change in its provider’s cross-platform deduplication method, so use the performance direction but not the withdrawn supply counts.

US averages hide concentration too: in a separate AirDNA revenue analysis covering April 2024–April 2025, the ten largest US markets accounted for 25% of modeled national STR revenue. That measures market revenue, not guest-nights or host profit, so it answers a different question from the seasonal shares above. It reinforces the need to check the actual destination.

Your next check: compare nearby properties by month, property type, and legal availability. Estimate sellable property nights—not regional guest-nights.

Listings are not the same as available homes

Statistics Canada counted 355,070 Canadian STR listings in 2023, using Airbnb and Vrbo data supplied by AirDNA. It estimated that 107,266 met its narrower “potential long-term dwelling” definition: entire units listed more than 180 days, excluding vacation-type properties. The agency explicitly cautions that this is not an observed count of homes removed from long-term housing. These older Canadian data illustrate why local property type and legal context matter; they are not a 2026 supply estimate.

Policy can change availability

A 2025 academic study of San Francisco’s Airbnb registration requirement estimated that, relative to surrounding untreated cities, availability fell 20–27%, booked nights fell 22–31%, and booking prices rose 3.3%. The study’s published abstract reports estimates for this policy and place; it does not make them a prediction for another city or a reason to dismiss local housing concerns.

Check your local authority’s current rules, permitted use, and night limits before using any market number to justify an investment.

Do not treat platform earnings as profit

Airbnb reports about €17.4 billion in EU host earnings for 2025. That is one platform’s self-reported figure—not the whole STR market and not net profit. Its “typical” host earned just over €4,600, but the report does not define “typical” or give the host denominator. Do not use it as an average or a forecast. It cannot be divided by Eurostat’s three-platform nights to calculate yield.

Use the worksheet instead: booked property nights × revenue actually received per booked night, minus per-stay, per-night, and fixed costs. Test a weak-season case too. Cleaning, utilities, fees, tax, repairs, financing, and your time are not erased by a market headline. This is planning arithmetic, not financial advice.

Stress-test the weak month

Worked example — invented US-dollar figures, not a market benchmark: suppose one property can legally and practically sell 20 nights this month. Revenue received is US$120 per booked night, after platform deductions; each stay requires US$45 of cleaning, each booked night US$10 of utilities and consumables, and the month has US$900 of fixed operating costs plus a US$100 allowance for the host’s time. Keep the stay count separate from nights: the weaker month has four stays across ten booked nights; the stronger one has five stays across 14 booked nights.

The same property and nightly receipt can move from a negative to a positive monthly result as bookings and turnover change. This is not net profit, cash flow, a forecast, or an observed host result: taxes, financing, capital replacement, extra fees, cancellations, and any costs not in the example still need separate treatment. Use your own records and check both the busy and weak months before committing money.

Airbnb’s 24-hour standard and a sub-hour study measure different things

Airbnb’s regular response-rate metric looks at new inquiries and reservation requests answered within 24 hours over the past 30 days. Airbnb says that response rate can affect a listing’s position in search results; it does not publish a ranking uplift for sub-hour replies. Superhost status uses a separate test: Airbnb’s criteria require a 90% response rate over the previous 365 days, assessed quarterly, alongside other standards. Neither number is a conversion rate. Airbnb also reports Superhosts’ median earnings were 60% higher than non-Superhosts’ over October 2023–October 2024. Superhosts meet several criteria, so that comparison does not isolate the effect of replying within 24 hours or establish an income lift from response speed.

The vendor’s 24-hour response-rate groups: IntelliHost reports average conversion of 0.5% for its 0–89% response-rate group, 0.8% for 90–99%, and 1.0% for 100%. Its article later describes the lowest group as below 89%, so the precise boundary is not fully consistent in the source. It does not define the conversion denominator or publish a controlled test. Read these as the vendor’s cohort observations—not Airbnb’s published conversion statistics.

The under-one-hour comparison: IntelliHost reports 1.0% conversion for listings replying in under an hour and 0.8% for those taking over an hour, a 0.2-percentage-point difference (25% relative to 0.8%). It also reports 16% higher daily impressions and a 3% increase in first-page impression rate for the faster group. Its response-rate comparison, by contrast, found no noteworthy difference in first-page ranking across response-rate groups. IntelliHost does not say whether the 3% is a relative or percentage-point change; it also gives no booking denominator or time window for the “up to 116% more bookings” comparison between below-89% and full response-rate groups. The study reports associations, not measured revenue or host-income effects, and cannot show that faster replies caused more bookings.

A peer-reviewed observational study of historical Airbnb shared-home listings also found an association between sub-hour responses and bookings across several regions. It supports the direction of the relationship, not IntelliHost’s precise percentages or a causal forecast. Airbnb also says more than half of trip requests are accepted within an hour after the host reads them; that is a platform statistic about request acceptance after reading, not proof that all guest messages receive a response within an hour.

Apply the evidence carefully: keep the 24-hour response-rate requirement intact, but do not treat it as a complete service measure. Track first-reply time, booking conversion and income separately on comparable listings and periods. Airbnb does not require a host to send the last message in a thread to maintain response rate; that policy does not mean an open guest need has been resolved. After booking, keep the owner and next step clear until access works or the promised repair is complete. For the fuller response evidence and workflow, see our guest-messaging guide. Native messaging plus a manual checklist may suffice for a small property; local judgement and physical work still need people.

Before committing money, write down the property’s address, permitted use, available nights, monthly local comparables, expected and weak-season booked nights, revenue received, all material costs, and the person responsible when a guest or cleaner needs physical help. Revisit the plan with observed bookings rather than repeating a platform-wide growth figure.

Use each figure for the question it measures: listing counts describe Airbnb supply; Inside Airbnb’s listing-nights are modelled from reviews; ONS and Eurostat report platform guest-nights in separate geographies; and AirDNA estimates US STR performance. They cannot be combined into one global demand rank. None predicts one property’s profit or proves a causal revenue benefit from faster replies. The worked budget is fictional, Airbnb’s earnings figure is platform-authored, and the San Francisco estimate concerns one local rule. Sources were checked on 30 September 2026.