Analyse instantanée gratuite
Révélez les revenus Airbnb pour n'importe quelle adresse ou ville
Performances Airbnb à San Francisco par nombre de chambres
Données Airbnb et Vrbo pour l'ensemble du marché de San Francisco.
Studio
Annonces actives
142
Performance inférieure
Revenu annuel
7,2 k $US
Prix par nuit
188,4 $US
Occupation
13 %
Rendement brut
91.6%
Performance type
Revenu annuel
35,0 k $US
Prix par nuit
206,6 $US
Occupation
45 %
Rendement brut
4.4%
Performance supérieure
Revenu annuel
52,5 k $US
Prix par nuit
254,2 $US
Occupation
52 %
Rendement brut
6.6%
1 chambre
Annonces actives
538
Performance inférieure
Revenu annuel
16,6 k $US
Prix par nuit
155,6 $US
Occupation
29 %
Rendement brut
2.1%
Performance type
Revenu annuel
60,3 k $US
Prix par nuit
207,2 $US
Occupation
67 %
Rendement brut
7.6%
Performance supérieure
Revenu annuel
105,0 k $US
Prix par nuit
348,6 $US
Occupation
71 %
Rendement brut
13.3%
2 chambres
Annonces actives
350
Performance inférieure
Revenu annuel
28,0 k $US
Prix par nuit
282,9 $US
Occupation
28 %
Rendement brut
2.1%
Performance type
Revenu annuel
86,0 k $US
Prix par nuit
333,6 $US
Occupation
62 %
Rendement brut
6.5%
Performance supérieure
Revenu annuel
178,5 k $US
Prix par nuit
609,7 $US
Occupation
68 %
Rendement brut
13.4%
3 chambres
Annonces actives
234
Performance inférieure
Revenu annuel
40,2 k $US
Prix par nuit
412,9 $US
Occupation
28 %
Rendement brut
2.1%
Performance type
Revenu annuel
120,8 k $US
Prix par nuit
522,4 $US
Occupation
57 %
Rendement brut
6.2%
Performance supérieure
Revenu annuel
255,3 k $US
Prix par nuit
820,0 $US
Occupation
70 %
Rendement brut
13.1%
4+ chambres
Annonces actives
97
Performance inférieure
Revenu annuel
41,1 k $US
Prix par nuit
614,4 $US
Occupation
19 %
Rendement brut
1.4%
Performance type
Revenu annuel
157,1 k $US
Prix par nuit
779,1 $US
Occupation
50 %
Rendement brut
5.3%
Performance supérieure
Revenu annuel
416,4 k $US
Prix par nuit
1,6 k $US
Occupation
63 %
Rendement brut
14.0%
| Chambres | Groupe de performance | Revenu annuel | Prix par nuit | Occupation | Rendement brut | Annonces actives |
|---|---|---|---|---|---|---|
| Studio | Performance inférieure | 7,2 k $US | 188,4 $US | 13 % | 91.6% | 142 |
Performance type | 35,0 k $US | 206,6 $US | 45 % | 4.4% | ||
Performance supérieure | 52,5 k $US | 254,2 $US | 52 % | 6.6% | ||
| 1 chambre | Performance inférieure | 16,6 k $US | 155,6 $US | 29 % | 2.1% | 538 |
Performance type | 60,3 k $US | 207,2 $US | 67 % | 7.6% | ||
Performance supérieure | 105,0 k $US | 348,6 $US | 71 % | 13.3% | ||
| 2 chambres | Performance inférieure | 28,0 k $US | 282,9 $US | 28 % | 2.1% | 350 |
Performance type | 86,0 k $US | 333,6 $US | 62 % | 6.5% | ||
Performance supérieure | 178,5 k $US | 609,7 $US | 68 % | 13.4% | ||
| 3 chambres | Performance inférieure | 40,2 k $US | 412,9 $US | 28 % | 2.1% | 234 |
Performance type | 120,8 k $US | 522,4 $US | 57 % | 6.2% | ||
Performance supérieure | 255,3 k $US | 820,0 $US | 70 % | 13.1% | ||
| 4+ chambres | Performance inférieure | 41,1 k $US | 614,4 $US | 19 % | 1.4% | 97 |
Performance type | 157,1 k $US | 779,1 $US | 50 % | 5.3% | ||
Performance supérieure | 416,4 k $US | 1,6 k $US | 63 % | 14.0% |
Les groupes de performance faible, type et élevée sont des références de marché, et non des résultats garantis. Données mises à jour en sept. 2026.
How much does a San Francisco Airbnb actually make, and can you run one somewhere you don't sleep?
Well, the median whole home inside the city took in $40,146 over its latest twelve months in BNBCalc listing data, but you can't earn that from a place you don't live in, and that condition rewrites every figure below it. The City and County of San Francisco certifies a short-term rental only where the host is the permanent resident, which the city measures as sleeping in that unit at least 275 nights a calendar year. Pull those nights out of the calendar and 90 remain, which is precisely the ceiling the city puts on stays where guests have the place to themselves, while nights when you're at home with them carry no limit at all. So if you pictured buying a condo and letting the whole thing out nightly from wherever you really live, unfortunately no amount of money makes that plan available here, and the closest thing the city allows is a furnished rental of more than 30 nights, a separate category that needs its own permit.
Live there yourself and it gets more interesting, although you'd be joining a thinner field than a year ago. Active listings across the whole market BNBCalc tracks here fell over the past year, with more than half of that drop arriving in a single step in September and October 2025, while average nightly rates climbed about 30% and occupancy hardly budged. Almost all of those listings sit inside San Francisco itself, where city and county are a single government, so every rule below is the city's own. The small remainder lies just across the county line in Daly City, in San Mateo County, where Daly City writes its own rules.
How Much Do San Francisco Airbnbs Earn in 2026?
That $40,146 is the median for a whole home inside the city, at a median nightly rate of $323, and the top quarter of listings started at $65,907. Neither figure is something a certified host can plan on, though, because the typical whole home in the city sold about 142 nights over the year and a certified host gets 90 unhosted ones. Spend all 90 at that $323 rate and the unhosted side of your year is worth roughly $29,000 before costs, and since the average stay here runs just under five nights, those 90 nights come to about 19 whole-home bookings.
So a San Francisco pro forma is mostly a room you rent out while you're home, with one short unhosted season laid on top wherever the rates are strongest. Make sure the model you build describes that business, because an annual revenue figure describes a different one.
Sort the market by bedroom count and the answer runs backwards from what you'd guess, at least on gross yield, which measures a year's revenue against what the home costs to buy. In BNBCalc's per-bedroom figures for this market it peaks on one-bedrooms and slides steadily as the homes get bigger, so three- and four-bedroom houses finish near the bottom, with studios below them again. Revenue does rise with size, only nowhere near as fast as the purchase price does. Then again, you're choosing a home you'll live in, so treat that ranking as a tiebreaker at most.
Management companies still have a role here, although the certificate stays with whoever sleeps in the unit. A manager can handle turnovers and guest messages for you, and the San Francisco property management roundup sets out the options, but remember that the registration and the 275 nights are yours to keep.
Is the San Francisco Airbnb Market Oversaturated?
Whoever ends up running the place, there's less competition around it than a year ago, although more than half of the change came in one step.
| Metric | Year-over-year change |
|---|---|
| Average nightly rate | +30% |
| Purchase price | +7% |
| Occupancy | +2% |
| Active supply | −33% |
BNBCalc data across all listings BNBCalc tracks in the market, setting September 2025 through August 2026 against September 2024 through August 2025, with every change rounded. Booking lead time is left out because this year's data can't show which way it moved, and any measure failing that test is dropped the same way. More than half of the fall in active supply came in one step in September and October 2025, the first two months of the window, and the data can't yet say why. Purchase price follows home prices rather than listings.
Probably not, although one unusual stretch makes that a lean. Active supply fell about 33% over the year, but more than half of the fall arrived in a single step in September and October 2025, when Los Angeles, Oakland and several other big-city markets dropped sharply too, and the data can't yet say whether hosts left or their listings were delisted. Occupancy, meanwhile, edged up about 2%, which is near enough to unchanged, and average nightly rates rose about 30%. In a crowded market you'd see hosts cutting prices to win bookings and still selling fewer nights, whereas here they charged about 30% more and sold the same share of their nights, which points away from crowding. Be careful with those two rows, though, because if the listings that vanished weren't typical of the rest, part of that rise in rate and occupancy may reflect which listings are left.
I wouldn't plan around that drop repeating, because it came as a single step and supply has held roughly level since. I wouldn't plan around it holding either, because hosts leaving, listings being delisted and the data picking up fewer of them would each leave you facing a different market a year from now, and nothing here tells them apart yet. Home values rose as well, about 7% over the same twelve months, so part of what you save on competition gets paid back at purchase. To see whether the crowd is returning, watch how full the listings you'd compete against are and check again every few months, because no headline figure will tell you honestly.
Booking lead time is left out of the table because its year-over-year change flips depending on which months you compare, but its level is steady enough to use. Over the latest twelve months a booking here arrived roughly 40 days before check-in on average, which is distant enough that an empty-looking calendar three weeks out isn't the emergency it would be somewhere last-minute.
When Is San Francisco's Peak Airbnb Season?
Those 40 days matter more once you see how unevenly the year is priced.
| Month | Occupancy | Avg nightly rate |
|---|---|---|
| Sep 2025 | 48% | $320 |
| Oct 2025 | 48% | $333 |
| Nov 2025 | 42% | $328 |
| Dec 2025 | 34% | $340 |
| Jan 2026 | 39% | $320 |
| Feb 2026 | 40% | $339 |
| Mar 2026 | 47% | $336 |
| Apr 2026 | 43% | $327 |
| May 2026 | 44% | $349 |
| Jun 2026 | 52% | $403 |
| Jul 2026 | 47% | $383 |
| Aug 2026 | 48% | $367 |
BNBCalc market data across all listings in the market, one row per month for September 2025 through August 2026. Occupancy and nightly rate are averaged independently, so read each column on its own.
San Francisco's peak season is summer, and it announces itself in the rate column far more than in how full the calendar gets. Between September and April the average nightly rate never moves more than about $25, holding in the low to mid $300s the whole way, with September and January the cheapest months at about $320 each. Then summer lifts it hard, to $349 in May and $403 in June, the highest of the twelve months, and it's still $367 come August. Price the year off its average rate, then, and you'll overcharge in the winter and undersell the summer.
Occupancy is the quieter column. December marks the year's floor at 34%, and from November through April only March sells above 43% of its nights, while every month from May through August fills at least 44%. Don't read much into gaps of a few points between those two stretches, though, because the months from May 2026 on come from a different load of BNBCalc's data than the months before, and occupancy can shift a few points across that join. The rate column carries no such doubt, and it's the one that tells you what summer is worth.
The week has a shape as well, though a surprisingly flat one. Saturday sells about a tenth more of its nights than the average day and Friday about 8% more, whereas Monday sells about 7% fewer. Rates barely react, with the two weekend nights about 5% over the average and every weekday within a couple of points under it, so there's little reason to discount your weekdays hard.
If you're the one living there, the two patterns hand you a plan. Those 90 unhosted nights are the scarcest thing you own, so spend them across June, July and August when the rate is at its best, and since bookings land about 40 days ahead on average, make sure your summer dates are open at summer rates by late April. Then let the spare room carry the winter, when a guest down the hall costs you nothing but company.
Where Should You Buy an Airbnb in San Francisco?
The spare room and the summer season work the same way anywhere in the city, but the rate you can charge for them changes from street to street. Nearly every listing in this market falls inside one of the city's 41 official analysis neighborhoods, so here are the 15 strongest, sorted by median annual revenue.
| Rank | Neighborhood | Median annual revenue | Median nightly rate | Median occupancy | 75th-percentile revenue |
|---|---|---|---|---|---|
| 1 | Financial District | $91,413 | $864 | 39% | $126,895 |
| 2 | Marina | $83,222 | $588 | 39% | $109,460 |
| 3 | Pacific Heights | $60,870 | $463 | 41% | $129,253 |
| 4 | Haight Ashbury | $58,853 | $467 | 41% | $81,887 |
| 5 | North Beach | $58,433 | $514 | 31% | $82,730 |
| 6 | Presidio Heights | $58,024 | $352 | 44% | $78,254 |
| 7 | Castro/Upper Market | $57,003 | $386 | 42% | $105,584 |
| 8 | Nob Hill | $56,938 | $540 | 37% | $107,820 |
| 9 | Twin Peaks | $54,818 | $454 | 42% | $60,147 |
| 10 | South of Market | $51,276 | $415 | 40% | $82,330 |
| 11 | Mission Bay | $50,391 | $574 | 30% | $70,352 |
| 12 | Noe Valley | $48,445 | $346 | 41% | $68,896 |
| 13 | Lone Mountain/USF | $47,555 | $418 | 41% | $117,285 |
| 14 | Potrero Hill | $47,316 | $334 | 42% | $78,351 |
| 15 | Russian Hill | $44,893 | $475 | 33% | $68,846 |
BNBCalc 2026 listing data for whole-home listings inside San Francisco's 41 official analysis neighborhood boundaries, trailing twelve months. Private and shared rooms are excluded, and so are listings that were rarely open or rarely booked. Revenue, rate and occupancy are medians, and the final column is the 75th percentile, the point where a neighborhood's top quarter of listings begins. Each column is worked out separately, so they don't multiply together. Neighborhoods below a minimum sample are left out, and rows close to that minimum are thin and can move between refreshes.
Every row in that ranking is a whole home, because private and shared rooms are filtered out before any figure is worked out, and that changes how you should read it. A neighborhood's median occupancy is the share of the year's 365 nights its middle listing sold, so the 37% to 44% most rows show works out to roughly 135 to 160 nights, and even Mission Bay's 30% comes to more than 100. Every row is past the 90 unhosted nights a certified host gets, and more than nine in ten of the city's whole-home listings sold more than 90 over the past year.
Part of that gap has a legal explanation. The city lets tourist hotels and approved timeshares rent all their units short-term without registering, and a stay of more than 30 nights falls outside the short-term rules, although fewer than one in a hundred of these listings average stays that long. You can't tell from the data which listings are certified, so read each row as what whole homes in that neighborhood earn today, never as what the rules would let you earn from one.
To turn a row into your own plan, multiply its median nightly rate by 90, which gives you roughly what a full unhosted season is worth there: about $35,000 at Castro/Upper Market's $386, for instance, or about $29,000 at the city's $323. Anything your plan needs beyond that has to come from renting a room while you're home, and keep in mind that this ranking can't price that for you, since rooms aren't in it.
Compare the occupancy figures with one another, and keep them away from the monthly seasonality table, which follows a broader set of listings across the whole market. Most rows land between 37% and 44%, while median nightly rates run from $334 to $864, so it's the rate column that sorts this ranking. North Beach, Mission Bay and Russian Hill break the pattern, all three filling noticeably fewer nights while charging more than most rows in the table, and no row owes its place to size, since every one of them has a one- or two-bedroom home in the middle.
Nob Hill and Castro/Upper Market rest on more listings than any other rows in the top fifteen, and they tell different stories. A large share of Nob Hill's listings sit with operators running many units each, the same pattern that fills the Financial District at the top of the ranking, where the middle listing is a one-bedroom going out at $864 a night. That's stock run as a business at scale, which a resident with one registered home can't copy, so I wouldn't take either row as a guide to what your own place would earn. Castro/Upper Market is the opposite case: most of its listings belong to hosts with a single unit, it holds 42% occupancy, among the highest in the top ten, and its $386 median rate is the one worth converting.
Most of the other rows rest on much thinner samples, Marina and the Financial District included, so treat those figures as leads to check. If your bet is on your own operating skill more than the address, Pacific Heights and Lone Mountain/USF carry unusually high 75th percentiles against their medians, and gaps that wide say a good operator there pulls clear of the crowd. Twin Peaks is the reverse, with its top quarter starting barely above its median, so there's little sign that running the place better moves the number much there.
Depth and revenue don't travel together, either, because the biggest concentration of listings anywhere in the city sits out in Sunset/Parkside, where the median lands at $36,122, below the city's $40,146 and well outside the fifteen shown.
The county line deserves one caution too, since a small slice of this market lies just south of it in Daly City, in San Mateo County, where the median whole home took in $46,716, above San Francisco's own. Those listings answer to Daly City's rulebook, so read BNBCalc's Daly City short-term rental guide before assuming a few blocks south buys you an easier ride.
The home you buy is also the home you'll live in, so the practical answer is to shortlist the neighborhoods you'd happily live in anyway and read their rows for rate, because rate is what your 90 unhosted nights are worth. For a closer look at individual neighborhoods, see the best San Francisco neighborhoods for Airbnb.
Which Amenities Make the Most Money in San Francisco?
Whichever street you settle on, the fit-out is the last thing to decide, and a handful of items move revenue more than anything else on the list.
Pooled across every bedroom count, BNBCalc's amenity model puts a barbecue at about 9% more revenue for a San Francisco listing. Yet what that pooled figure hides is how much the leading feature shifts once you sort by bedroom count. A barbecue heads the one-bedrooms too, but a gym is worth about 20% on a two-bedroom and a hot tub about 17% on a three-bedroom, a swing big enough that fitting out a particular unit off the market-wide figure would send you shopping for the wrong thing.
Keep in mind that each figure compares listings that have the feature with similar listings that don't, which isn't the same as a return on buying it, and a gym usually comes with the building, so it's a reason to favor one building over another more than something you'd install. Letting guests bring pets and offering parking carry a measured lift in San Francisco too, though BNBCalc Markets keeps their percentages behind its paywall.
Then there are cleaning fees, which aren't the profit center hosts take them for. As of September 2026 about 59% of whole-home listings across the market carried one, averaging around $135 on the homes that charged. Spread over every listing, including those charging nothing, it works out near $2,850 a year, and virtually every dollar of that leaves again with whoever does the cleaning. So the honest way to read a cleaning fee is as the turnover billed at cost.
Is Airbnb Legal in San Francisco?
Yes, in the home you sleep in and nowhere else, which settles the question for anyone hoping to buy a place and rent it out.
Any stay under 30 nights is a short-term rental in the city's eyes, as of September 2026, and doing one legally takes two approvals in sequence. First comes a Business Registration Certificate from the Office of the Treasurer and Tax Collector, which you need in hand before you apply to the Office of Short-Term Rentals. Then comes that office's certificate, whose number has to appear on every listing you post. The application fee is $925, you don't get it back whether you're approved or refused, and the certificate lasts two years.
Qualifying rests on the permanent residency test. You have to occupy the unit at least 275 nights of the calendar year you rent it, or 75% of the days you've owned or rented it if this is a first partial year, and 60 consecutive nights of living there have to be behind you before you apply. Unhosted stays of 30 nights or more don't get you around that count, either, since the city weighs them when it checks your 275. Owners and tenants can both register, although the city writes to the owner of record whenever a tenant applies, and a tenant in a rent-controlled unit can't take in more from short-term guests in any month than the rent they pay that month. You get one unit, so even in a building you own outright it's still only the one you sleep in.
Three further conditions catch people out. You need at least $500,000 of liability cover for the rental use, naming the building's owners on the policy, unless every booking runs through a platform whose coverage matches or beats it. Hosted and unhosted offerings need separate listings, because they're separate products to the city and the quarterly report asks about both. And a fair number of homes can't be registered at all: the Presidio itself (a different place from the Presidio Heights neighborhood), Fort Mason and Treasure Island; single-room occupancy units and dorms; below-market-rate and public housing; accessory dwelling units; anything hit by an Ellis Act eviction after November 2014; and anywhere guests would sleep outdoors or in a detached structure. Check the city's Property Information Map for your zoning, too, because in an RH-1(D) district notice goes out to everyone within 300 feet and the application sits 45 days while neighbors comment.
Approval then starts a rhythm that never stops. Quarters open on January 1, April 1, July 1 and October 1, and 30 days after each one closes you report the nights you rented, hosted and unhosted, filing a zero when nothing happened. Business registration renews annually. The certificate renews every two years, and that renewal asks you to show 275 nights of occupancy in each of the two preceding calendar years, which is why the record-keeping isn't optional theater.
Tax lands on the guest's invoice instead of coming off your rate, although answering for it is still your job. San Francisco charges a 14% transient occupancy tax on stays under 30 nights, worked out on the listing price with the cleaning fee included, and above that sits the Tourism Improvement District assessment at 2.25% in Zone 1 or 2.00% in Zone 2. California adds no state lodging tax, so both charges belong to the city. Airbnb is one of three companies the Tax Collector has approved to collect and remit both charges for hosts, which leaves you nothing to file if you book exclusively through it, whereas a Vrbo or direct booking means your own return by January 31. For the state side of the filing, see the California guide to short-term rental taxes.
The city enforces through the platforms as well as against hosts. A hosting platform may only take a booking, or charge for one, after taking reasonable care to confirm the unit is lawfully registered, and a listing showing the wrong registration number can be pulled and its pending reservations cancelled. On the host's side, the city's short-term rental office quotes penalties of $484 a day per unit, accruing from the day a notice of violation is issued until the activity stops. That figure comes from the Administrative Code, which caps a first violation at four times its $121 hourly administrative rate for each unit and each day, and a second or later one at eight times, or $968 a day. Repeat violations pull the unit off the registry for a year and onto a published list, and the city can escalate to civil penalties of up to $1,000 a day or a misdemeanor charge. Every form, document, tax account and contact number is collected in the San Francisco short-term rental regulation guide.
Where Do These San Francisco Airbnb Numbers Come From?
Paperwork lives over in that guide. The market measurements in this one, from the year-over-year changes to the monthly rates, were taken from BNBCalc Markets, the research half of BNBCalc, where short-term rental markets get lined up side by side so a city can be chosen on evidence. Since a San Francisco year turns on price far more than on nights sold, seasonality is the panel I'd pull up there, because it shows where a market's trade bunches across the calendar. That's the quickest way to find out whether the shape you've just read repeats anywhere else on your list.
How Do You Estimate Airbnb Revenue for a San Francisco Property?
A city-wide median only takes you so far once there's one address in front of you. Begin at the San Francisco market page, because revenue, occupancy and the year's shape all sit there in their current form. If the city itself is still undecided, put it next to the rest of the state on California's markets ranked by gross yield. With a particular home in front of you, run its real price, loan and operating costs through BNBCalc and read the return it gives you.
Then test that answer against three local realities. Whether you'll sleep there 275 nights is the first, because if you won't, the nightly figures are irrelevant and the place should be priced as a furnished rental of more than 30 nights, with the permit that takes. The 90-night unhosted ceiling is the second, so model a spare room running most of the year with a short whole-home season on top, never a whole home running all year. Third, price with the 14% and the district assessment in view, since they're added to what the guest pays over your rate and cleaning fee, and they move what a guest is willing to agree to.
Anywhere a city decides who may host before it decides anything else, the averages only ever describe whoever got through the door. That makes eligibility the first thing to check about a market, ahead of any number it publishes.
Frequently Asked Questions
What Is the Average Airbnb Income in San Francisco?
In BNBCalc listing data, the median whole home inside the city made $40,146 across its most recent twelve months in 2026, at a median nightly rate of $323, and the top quarter of listings started at $65,907. Those are medians, meaning the middle listing, and they describe homes rented whole across the year. A certified host is capped at 90 unhosted nights, worth roughly $29,000 at that median rate before costs.
Is Airbnb Still Profitable in San Francisco in 2026?
It can be, though only for hosts who qualify, since San Francisco certifies short-term rentals solely where the host is the permanent resident. The latest year brought them less competition and higher prices: active supply fell about 33% across the market, although more than half of that came in a single step in September and October 2025, alongside similar drops in Los Angeles and other big-city markets, and the data can't yet say whether hosts left or listings were delisted. Average nightly rates rose about 30% and occupancy about 2% across those twelve months. Whether a specific home profits still turns on its price and running costs.
What Is the Best Month for Airbnb in San Francisco?
June 2026 was the dearest month of the twelve, averaging $403 a night, while December 2025 hit the occupancy floor at 34%. Rates hold in the low to mid $300s from September through April, then jump to $349 in May and peak at $403 in June, easing back to $367 by August. Friday and Saturday fill slightly more of their nights than other days, although their rates sit only about 5% above the average.
Do You Need a Permit to Run an Airbnb in San Francisco?
Yes, and two of them. Hosting any stay under 30 nights takes a Business Registration Certificate from the Office of the Treasurer and Tax Collector, obtained first, then a certificate from the Office of Short-Term Rentals, whose number must appear on every listing. The application fee is $925, it's non-refundable, and approval lasts two years. Hosts also need $500,000 of liability coverage unless their platform supplies it.
Can You Buy an Investment Property in San Francisco and Run It as an Airbnb?
Not as a short-term rental. San Francisco certifies short-term rentals only where the host is the unit's permanent resident, occupying it at least 275 nights a calendar year, and each host gets a single unit. Platforms have to take reasonable care to confirm a unit is registered before booking it. A home you don't live in can still be rented furnished for more than 30 nights at a time, which the city treats as an intermediate-length occupancy needing its own permit, and tenant protections and rent control may apply to it.
How Many Nights Can You Rent an Airbnb in San Francisco?
Unhosted stays, where guests have the unit to themselves, are capped at 90 nights per calendar year. Hosted stays, where you're in the unit overnight with the guest, carry no annual limit. The 90 follows from the duty to occupy the home 275 nights a year, which leaves 90 over, and unhosted stays of 30 nights or more also count when the city checks that 275. Hosted and unhosted offerings need separate listings, and both get reported to the city every quarter.
Which San Francisco Neighborhood Is Best for Short-Term Rentals?
It depends what you're buying, and since you'll live there, a neighborhood's nightly rate matters more than its annual median. The Financial District posts the highest median revenue, $91,413, but on a thin sample with a large share run by operators holding many units. Castro/Upper Market is the deepest row in the top fifteen run mostly by single-unit hosts, with a $57,003 median and a $386 median rate, which puts a full 90-night unhosted season there at about $35,000. Each figure is a 2026 median from BNBCalc's whole-home listing data, grouped by the 41 neighborhood boundaries the city publishes.
How Much Is the Airbnb Tax in San Francisco?
Any San Francisco stay under 30 nights carries 14% transient occupancy tax for the guest, charged on the listing price with the cleaning fee included, plus a Tourism Improvement District assessment of 2.25% in Zone 1 or 2.00% in Zone 2. California levies no state lodging tax above that. Airbnb is approved to collect and remit both for hosts, while a Vrbo or direct booking leaves the host filing annually by January 31.
Airbnb Tax Deduction Calculator
Paying too much in taxes? We have the perfect solution. Simulate an Airbnb home purchase below.
Purchase Price
$450K
Structure Value
70%
Apply Trump's Tax Cut (Bonus Depreciation)
Depreciation
$117,695
Interest
$21,600
Tax
$6,750
Year 1 Deduction
$146,045
Want to claim this deduction? Get a free cost segregation benefit analysis from CSA Partners — no obligation.
Get Full Analysis

