California permitted 198,596 accessory dwelling units between 2013 and 2025 across 475 cities — and 2025 set a new all-time high at 37,423 city-issued permits (41,831 including county-unincorporated areas). This analysis uses California HCD's Annual Progress Report data released on September 4, 2026, which carries the complete 2025 reporting year for the first time; we re-ran the dataset to replace the partial 2025 figure we published in May. Headlines: 2025 is the new peak, but growth has slowed to +5.5% (from +67% in 2019 and +11% in 2024); Los Angeles grew faster than the state and its share ticked up for the first time since 2018; and San Diego, the state's second market, recorded its first year-on-year decline. Raw data is downloadable below.
California's modern ADU era began on January 1, 2017, when SB 1069 and AB 2299 took effect and stripped cities of much of their discretionary review authority. Statewide permits jumped from 1,336 in 2016 to 5,154 in 2017 — a near-quadrupling in twelve months. The 2017 reforms were designed to be hard for cities to resist; on this measure they worked.
The 2017 jump was followed by a second, larger one in 2019 when SB 13, AB 68, and AB 881 took effect. That bundle extended state preemption further: it banned owner-occupancy requirements, capped impact fees, mandated 60-day ministerial review, and required cities to allow JADUs alongside ADUs. Permits jumped from 9,000 in 2018 to 15,064 in 2019 — a 67 percent year-over-year increase, the largest growth rate of the post-preemption era.
The COVID year was nearly flat (16,024 permits in 2020, +6%), then the recovery bent sharply upward. 2021 came in at 24,627 (+54%), 2022 at 31,814 (+29%). Through the Big West Coast Dataset's 2022 endpoint, the curve still looked steep.
HCD's data shows the curve has decisively bent. 2023 came in at 35,723 statewide (+12%), 2024 at 39,633 (+11%), and 2025 at 41,831 (+5.5%) — a new all-time high, but the smallest annual gain since the 2017 reforms. Growth has compressed from the +50–67% bursts that followed each preemption package to mid-single digits. The market is maturing: 40,000-plus permits a year is a meaningful share of California's housing supply, but the explosive-growth phase of 2017–2022 is over.
Two things changed between our May 2026 version of this analysis and this one. First, the 2025 figure went from a partial 25,289 (360 jurisdictions had filed by HCD's 1 April deadline) to 41,831 (467 jurisdictions) once late filers were in — a reminder that HCD's first snapshot of any reporting year understates it by a third or more. Second, earlier years were revised as cities amended their filings: 2024 rose from 38,910 to 39,633 (+1.9%), and 2018–2023 each moved by less than 0.3%. Every figure here now tracks HCD's September 2026 release.
Through 2022, Los Angeles alone permitted roughly 35% of all city-issued California ADU permits. On the full 2013–2025 window its lifetime share is 36.0% (71,489 permits) — and, for the first time since its 2017–2018 peak, its share of the latest year rose: 32.9% of 2025 city permits, up from 32.1% in 2024, because LA grew 8% (11,412 to 12,327) while the rest of the state grew about 4%.
San Diego remains the second mass-volume market at 9,250 lifetime permits (4.7% of the state), but 2025 brought its first decline of the HCD era: 2,165 permits, down 5% from 2,282. San Jose (3,521 lifetime, 502 in 2025), Long Beach (3,493 / 644) and Glendale (3,465 / 695) round out the top five; Glendale was the only one of the three to grow in 2025 (+24%).
Bay Area performance remains weak. San Francisco has permitted 2,875 ADUs over thirteen years and slipped further in 2025 (171, down from 231 — its lowest year since 2018). Oakland permitted 192. The pattern holds: dense single-family cities (LA, Long Beach, Garden Grove, Glendale, Fullerton, Santa Maria) build ADUs at scale; high-rise and townhome cities struggle to convert state preemption into permit volume because the housing stock isn't garage-and-backyard.
The fastest movers in 2025 were mid-sized Southern California cities: Lakewood more than doubled (197 to 498), Fullerton rose 39% (424 to 589), Oxnard 29% (386 to 497), Glendale 24%. The largest declines were Burbank (411 to 241), Sacramento (533 to 378), San Diego, Lancaster (320 to 209) and Long Beach (747 to 644). The latest-year column in the table below shows the 2025 count for each of the top 15 cities.
| # | City | Lifetime 2013–2025 | % of CA | 2025 alone | Share (vs LA) |
|---|---|---|---|---|---|
| 1 | Los Angeles | 71,489 | 36.0% | 12,327 | |
| 2 | San Diego | 9,250 | 4.7% | 2,165 | |
| 3 | San Jose | 3,521 | 1.8% | 502 | |
| 4 | Long Beach | 3,493 | 1.8% | 644 | |
| 5 | Glendale | 3,465 | 1.7% | 695 | |
| 6 | San Francisco | 2,875 | 1.4% | 171 | |
| 7 | Garden Grove | 2,637 | 1.3% | 434 | |
| 8 | Santa Maria | 2,556 | 1.3% | 479 | |
| 9 | Fullerton | 2,303 | 1.2% | 589 | |
| 10 | Oakland | 2,108 | 1.1% | 192 | |
| 11 | Sacramento | 1,942 | 1.0% | 378 | |
| 12 | Anaheim | 1,877 | 0.9% | 441 | |
| 13 | Burbank | 1,783 | 0.9% | 241 | |
| 14 | Oxnard | 1,309 | 0.7% | 497 | |
| 15 | Campbell | 1,190 | 0.6% | 340 |
UC Berkeley's Center for Community Innovation graded California ADU ordinances in 2020. They evaluated 424 jurisdictions across criteria including parking requirements, owner-occupancy rules, fee structures, design review, and clarity of submittal requirements. Cities received letter grades A through F.
Cross-referencing those Scorecard grades against permit volume — including the 2023–2025 data — produces the same nuanced result the 2022-cutoff analysis showed. A-grade cities permit more on average than D or F-grade cities, but the average is dragged upward by Los Angeles. Removing LA, the median A-grade city permits more like the median B or C city than the gap suggests on paper. The Scorecard captures whether a city has gotten out of homeowners' way procedurally, not whether homeowners want to build.
A small city with a perfect ordinance and few aging-in-place homeowners willing to add backyard rentals will have a high grade and low permit volume. A larger city with a moderately permissive ordinance, severe housing pressure, and lots of convertible garages will have a B or C grade and high permit volume. The grade is necessary — the cities that block ADUs through procedural friction can effectively zero-out adoption — but it is not sufficient.
The Scorecard predicts whether a city has built the off-ramp. It does not predict whether anyone is on the highway.
Of the 475 California cities in our merged dataset, 72 — about one in seven — permitted fewer than ten ADUs across the entire 2013–2025 window. 19 cities permitted exactly zero. 166 cities (35 percent of the sample) permitted fewer than fifty over thirteen years.
The long tail is shrinking, slowly. The 2022-cutoff analysis had 137 cities under ten and 39 at zero; our 2024-cutoff version had 86 and 21. But the roughly 10,000-permit gain from 2022 to 2025 statewide came overwhelmingly from cities already in the top quartile growing further, not from the long tail starting to participate. 414 cities issued at least one ADU permit in 2025, and the top 25 still account for 62% of all city-issued permits.
Three hypotheses still explain the long tail. First, housing-stock incompatibility: many small Central Valley and Inland Empire cities have older housing without garages or basements suitable for conversion, and lot sizes that make detached new-build economically marginal at construction costs of $300+ per square foot. Second, demographic mismatch: ADUs are disproportionately built by homeowners aged 55–75 who want a rental income stream or housing for a family member. Third, post-preemption knowledge gaps: many small-city homeowners simply do not know that state law has overridden their city's restrictive 2015-era ADU ordinance, and city planning departments — particularly those that opposed the preemption legislatively — have little incentive to publicise the new rules.
The third explanation is the only one a state housing policy can address. HCD's expanded technical-assistance program (launched 2023) and the AB 1332 pre-approved-plan framework (2025) both aim at it. 2025 was the first full year under AB 1332 and the long tail did not visibly activate; the 2026 cohort of permits is the next test of whether ADUs become a statewide tool or remain an LA-plus-coastal-metros phenomenon.
The market has clearly entered its second phase. The first phase (2017–2022) was the ramp: state preemption created legal space, homeowners and builders learned the rules, and the curve compounded at +20–67% per year. The second phase (2023+) is maturation: +12%, then +11%, then +5.5% — a decelerating single-digit trend, modest expansion of the producing-cities base, and a slowly diversifying mix of project types (detached new-builds, garage conversions, prefabs, JADUs).
The 2022-era prediction in this article's first version — "if 2021–2022 trend continues, 2023 will likely cross 30,000 and 2026 should approach 50,000" — was right on 2023 and now looks out of reach for 2026. Our May 2026 version, written on partial data, projected 2026 landing "closer to 47,000". With 2025 complete at 41,831 and growth at +5.5%, a repeat of that rate lands 2026 near 44,000; even a return to 2024's +11% only reaches about 46,500. Still record territory, but the curve is flattening.
Three legislative changes could re-accelerate growth. AB 1033 (ADUs separately saleable as condos in opt-in cities) is now in effect with Berkeley, SF, San Diego, San Jose, and Oakland opted in or advancing — this creates an entirely new sub-market the historical data doesn't capture yet. AB 1332 (pre-approved ADU plans, in force since 2025) compresses permit timelines from 60 days to roughly 30 in cities like Burbank, Sacramento, and San Diego. SB 1211 (multi-family ADUs, expanded allowances) opens up a category most cities have under-utilised.
What the next dataset will not show, but that homeowners should watch, is the ratio of detached new-builds to garage conversions. The economics differ radically — a $300/sqft detached new-build at 750 sqft is a $225,000 capital project; a $150/sqft garage conversion at 400 sqft is $60,000. As construction costs continue to outpace wage growth, the conversion share will likely rise, and the absolute permit count may grow more slowly than the project-count would otherwise predict. HCD's APR data captures both as "ADU" — the granular split has to come from individual city permit records.
The biggest single-city question for 2026–2027 is whether San Diego's 2025 dip (2,165, down 5%) is noise or a plateau. San Diego permitted 2,282 in 2024 — a fifth of LA's number despite being a fraction of LA's population — on the strength of its bonus-ADU programme, the only programme nationally generating four-digit annual ADU volume from a non-LA city. 2026 will tell whether the programme keeps scaling or whether 2025 marked its ceiling.
Permit data 2013–2017: Big West Coast ADU Dataset published by BuildinganADU.com (Kol Peterson), drawn from city building-department records aggregated and normalised by Karen Chapple's research team at the UC Berkeley Center for Community Innovation. The dataset's January 2024 release covers 2013–2022; we use it for 2013–2017 because HCD's APR framework didn't begin systematic ADU-line-item tracking until 2018.
Permit data 2018–2025: California HCD Housing Element Annual Progress Report (APR) Table A2, downloaded from data.ca.gov on 2026-09-09. Last refreshed by HCD on September 4, 2026. Filter: UNIT_CAT='ADU' AND BP_ISSUE_DT1 non-empty (building permit was actually issued, not just entitled). We sum the NO_BUILDING_PERMITS field per row, treating empty/0 as 1 since the permit-issue date implies a permit was issued. This matches HCD's own methodology in its public-facing APR Dashboard.
Why the source mix: our first published numbers for 2018–2022 used the Big West Coast Dataset's January 2024 snapshot. HCD's data has been continuously refreshed since then as cities re-file with corrected counts; the 2022 figure in our first version (24,851) is 31,814 in HCD's current release — a 28% upward revision largely from late and corrected filings. We migrated 2018+ to HCD-current values in May 2026 so the methodology is one consistent source for all post-preemption years, and we refresh it each quarter.
2025 completeness and revisions: 467 jurisdictions reported 2025 ADU permits in HCD's September 2026 release — more than the 461 that reported for 2024 — so we treat 2025 as a complete year and no longer flag it as partial. HCD data is revised continuously as cities amend filings (2024 rose 1.9% between the April and September releases), so treat every figure as the September 2026 state of the record rather than a final count.
Ordinance grades: California ADU Scorecard 2020, UC Berkeley Center for Community Innovation, Appendix C. Grades reflect the legal text of each city's adopted ADU ordinance as of the 2020 evaluation. Subsequent state legislation (AB 2221, AB 976, AB 1033, AB 1332) raised the floor for every city; cities that have not updated their ordinances to reflect post-2020 state law may have effectively higher or lower grades than the Scorecard indicates depending on how much the local text now contradicts state law.
City vs county: we exclude county jurisdictions (LA County unincorporated, Alameda County unincorporated, etc.) from the city-level analysis above (top 15, long-tail counts, grade-vs-volume) because their permit-issuing geographies overlap with incorporated cities and double-count when both report. Including the 57 county jurisdictions brings the statewide 2013–2025 total from 198,596 (cities only) to 223,072.
Reproducibility: all numbers in this article are computed live at build time from data/permits_ca.csv (5,265 rows of city × year × permits). The full file is downloadable. The refresh script (scripts/refresh_permits_hcd.py) rebuilds the file from HCD's Table A2 in one step — the analysis is reproducible end-to-end.
The full extracted dataset is downloadable as CSV — 5,265 rows of city × year × permit count, covering 475 California cities and 57 county jurisdictions, 2013–2025. Cite as: "BackyardADU.net California ADU permits dataset, September 2026 release. Sources: Big West Coast ADU Dataset (2013–2017) + California HCD Annual Progress Report (2018–2025)."
If you're a journalist working on a California housing story and want underlying methodology notes, custom slices, or a specific city deep-dive, email hello@decisionlab.com.au. We're happy to help.
223,072 across all jurisdictions 2013–2025 (198,596 if you exclude county-unincorporated areas to avoid double-counting). 2025 alone added 41,831. This is the first public city-level analysis of HCD's September 2026 release, which carries the complete 2025 reporting year.
Los Angeles, by an enormous margin — 71,489 ADUs from 2013 to 2025, 36% of all city-issued California ADU permits in that window. The next-largest city, San Diego, permitted 9,250. LA permitted 12,327 in 2025 alone, up 8% on 2024.
Yes, on the metric they were designed to move. Statewide ADU permits jumped from 1,336 in 2016 to 5,154 in 2017 — a near-quadrupling — when SB 1069 and AB 2299 took effect. The 2019 follow-up package (SB 13, AB 68, AB 881) drove a further 67% year-over-year jump in 2019. By 2025 California was permitting 41,831 ADUs annually, a 31× increase over the 2016 baseline.
2025 — 41,831 statewide ADU permits issued (cities + county-unincorporated combined). That is a new all-time high, up 5.5% from 2024's 39,633 and 31% from 2022. Every year since 2017 has set a new record, but 2025's gain was the smallest since the reforms began.
Yes, materially. Annual growth was +67% in 2019, +54% in 2021, +29% in 2022, then +12% in 2023, +11% in 2024 and +5.5% in 2025. The market is in its mature phase. The 2017 and 2019 preemption-jump patterns won't repeat without new legislative shocks; the next year-over-year inflection will likely be driven by AB 1033 (saleable ADUs) or interest-rate moves rather than new state preemption.
Three reasons. First, housing-stock incompatibility — many older or smaller cities lack convertible garages or have lot sizes too small for detached new-builds at $300+/sqft construction costs. Second, demographic mismatch — ADUs are disproportionately built by homeowners aged 55–75. Third, post-preemption knowledge gaps — many homeowners and city planning departments have not yet adjusted to the fact that state law overrides previously-restrictive local ordinances. 72 of 475 California cities permitted fewer than 10 ADUs over the entire 2013–2025 window.
From /data/ — we publish the merged dataset as a 5,265-row CSV, free, CC BY 4.0 for any reuse with attribution. Original underlying sources: Big West Coast ADU Dataset (2013–2017) and CA HCD APR Table A2 (2018–2025, September 2026 release).