San Francisco guide · City & County, 27 ZIPs
House hacking San Francisco: 27 ZIPs, ranked.
We tier-ranked all 27 in-scope San Francisco ZIPs for one specific buyer: a house hacker with a 3–10% down payment who plans to live in the property. The ranking scores engineered growth (adopted upzones, funded transit, capital actually flowing), not vibes. Twelve ZIPs earned a full page; the other 15 didn't, and that's the point.
The three things that decide everything here
Housing stock decides more than capital here. Several ZIPs carry S- or A-strength capital flows and still land at C: the downtown conversion core, Mission Bay, Transbay, and Treasure Island all build high-rise condo, office-conversion, or new-construction BMR product with no fee-simple 2–4-unit or house-hackable single-family stock behind it. Every ZIP that ranks well pairs its catalysts with separable flats, houses over garages, or in-law potential you can actually buy.
The Family Zoning Plan is the biggest per-ZIP tier determinant. Effective Jan 12, 2026, it upzones roughly 96,000 parcels, mostly on the western and northern side. Parcel-level reach was confirmed in only two ZIPs at synthesis (94122's Judah corridor at 40→85 ft and the Outer Richmond); Glen Park's 140 ft BART lots and the Taraval corridor's 85 ft were confirmed at the July 14 re-verification. Everywhere else the corridor heights are still WARN: never underwrite an unconfirmed upzone.
The rent side does unusually heavy lifting, and rent control is real. AI-sector in-migration has rents compounding at ~22% a year, with 1-beds clearing $4,000 and 2-beds the highest in the nation, against fewer than 1,000 market-rate starts in the trailing 12 months. The same demand that inflates purchase competition services the mortgage. But on older multi-unit stock San Francisco rent control binds: verify vacancy-decontrol status per building (the Mission and the Western Addition carry the deepest exposure) before modeling market rents.
The twelve ZIPs worth your attention
| # | ZIP | Area | Tier | Median | The one-line case |
|---|---|---|---|---|---|
| 1 | 94124 | Bayview / Hunters Point / Candlestick | S | ~$933K | 10,672 planned units on the lowest price band in the city |
| 2 | 94107 | Potrero Hill / Dogpatch / Pier 70 | S | ~$969K–1.14M | Mission Rock delivering next to separable Victorian flats |
| 3 | 94112 | Excelsior / Outer Mission / Ingleside | A | $1,182,000 | The cleanest transit-oriented affordability play in the city |
| 4 | 94110 | Mission / Bernal Heights | A | $1,225,000 | Two BART stops and the deepest 2–4-unit stock in SF |
| 5 | 94134 | Visitacion Valley / Portola / Sunnydale | A | ~$898K | Lowest basis citywide with a delivering catalyst |
| 6 | 94122 | Inner/Outer Sunset / Golden Gate Heights | A | $1,399,000 | Confirmed 40→85 ft upzoning beside the $4.3B UCSF build |
| 7 | 94121 | Outer Richmond / Sea Cliff | A | $1,697,000 | 562 units filed at the Safeway site, plus a VA anchor |
| 8 | 94117 | Haight-Ashbury / Cole Valley / Ashbury Heights | A | ~$1,465,000 | Walk-to-UCSF Victorians beside an active $4.3B build |
| 9 | 94118 | Inner Richmond / Laurel Heights / Presidio Heights | A | $2,014,000 | 3333 California plus Geary BRT, at a $2M basis |
| 10 | 94115 | Western Addition / Fillmore / Japantown | A | ~$1,384,000 | A $2B redevelopment on the Geary corridor, long-horizon |
| 13 | 94131 | Glen Park / Diamond Heights / Twin Peaks | A | $1,627,000 | A confirmed 140 ft upzone at Glen Park BART |
| 17 | 94116 | Parkside / West Portal fringe | A | $1,610,500 | Delivered Taraval rail plus a confirmed 85 ft corridor upzone |
Tiers: S = multiple converging catalysts with capital already flowing; A = strong catalysts with less capital certainty. Ranks are the original priority order; 94131 and 94116 entered the A tier at the 2026-07-14 re-verification. The remaining 15 ZIPs graded B (7) or C (8): stock mismatches where real capital builds product a house hacker can't buy (FiDi, Transbay, Mission Bay, Treasure Island), preservation- and exclusion-capped cores (Chinatown, the Tenderloin's FZP carve-out, North Beach), premium corridor-only cases, and organic-only growth. 94143 is UC campus land with no fee-simple inventory at all. Rankings re-verified 2026-07-16.
Buy now or wait?
Act now on the outer and value ZIPs; stay price-disciplined in the core. Mid-2026 is the tightest SF for-sale market since 2021: 1.1 months of single-family supply, ~12 days on market, ~25% over asking, fewer than 700 active listings citywide. PMMS sits at 6.43% (Freddie Mac PMMS, a seven-week low; 6.67% a year ago) while rents compound at ~22% a year. And if rates ease another 25–50 bps into year-end as forecast, sidelined AI-cohort buyers push prices another leg higher against a structurally supply-constrained base.
The call is uneven. The buy-now case concentrates in Bayview, Visitacion Valley, the Excelsior, and the Outer Richmond and Sunset: $900K–$1.7M entries that pencil against $4,000+ 1-bed rents, with real negotiating room. In the core and premium ZIPs (the Haight, Inner Richmond, Castro, West Portal) you'd be overpaying at 25%-over-asking clearing prices: buy only where the specific building's rent stack clears the payment, and don't chase. Treasure Island and Mission Bay get underwritten unit-by-unit if at all; ground-lease, HOA, and BMR structure dominates that math.
Property taxes
San Francisco is a consolidated city and county, so every in-scope ZIP shares one rate: 1.183% combined effective (SF Treasurer FY2025-26, consolidated city and county); new buyers pay about 1.183% of purchase price in year one, then assessed value rises at the 2%/yr Prop 13 cap until resale.
The combined rate has held in the 1.17–1.19% band for five years, and no pending measure pushes it above ~1.20%. First-year tax runs about 1.183% of the purchase price (real money at SF prices: model it precisely), then Prop 13 caps assessed-value growth at 2% a year until resale.