Can I build an ADU? Not a yes

In California, the objections you are bracing for are prohibited by statute

Most people arrive at a feasibility question expecting a fight: parking, a hearing, a neighbor, a rule that you have to live on site. In California the Legislature has taken those away from the agency, in terms, and the reflex fear is now the expensive part — because it stops people asking at all.

Cite it correctly first. SB 477 (Stats. 2024, ch. 7), an urgency measure effective 25 March 2024, recodified State ADU Law out of Government Code § 65852.2 into Chapter 13 of Division 1 of Title 7, sections 66310 through 66342.

A guide that still cites § 65852.2 is citing a repealed numbering. State ADU law now sits at Gov. Code §§ 66310–66342.

That is a fast, free test of any guidance you are handed: a consultant deck, a blog or a counter handout still working from § 65852.2 has not been revised since March 2024, and the article has been amended twice since.

Replacement parking cannot be required. Gov. Code § 66314(d)(11): “When a garage, carport, covered parking structure, or uncovered parking space is demolished in conjunction with the construction of an accessory dwelling unit or converted to an accessory dwelling unit, the local agency shall not require that those offstreet parking spaces be replaced.” Where parking is required at all, Gov. Code § 66314(d)(10)(A) caps it: “Parking requirements for accessory dwelling units shall not exceed one parking space per accessory dwelling unit or per bedroom, whichever is less. These spaces may be provided as tandem parking on a driveway.” Under § 66314(e) a demolition permit for a detached garage being replaced by an ADU is reviewed with the ADU application and issued at the same time, and under § 66314(f) no written notice or placard is required for that demolition unless the property is in a historic district.

This is the single most common objection people brace for, and in California it is not available to the agency. It cannot make you replace the parking the garage was providing.

Nor an owner-occupancy requirement. Gov. Code § 66315, added by SB 477 (Stats. 2024, ch. 7), effective 25 March 2024, in full: “Section 66314 establishes the maximum standards that a local agency shall use to evaluate a proposed accessory dwelling unit on a lot that includes a proposed or existing single-family dwelling. No additional standards, other than those provided in Section 66314, shall be used or imposed, including an owner-occupant requirement, except that a local agency may require that the property may be used for rentals of terms 30 days or longer.” Read the shape of that sentence, not just its subject.

It is not a list of things a city may not do; it is a ceiling on the whole evaluation. Section 66314 is the maximum standard, and no additional standard may be used or imposed.

The owner-occupancy ban is named inside that ceiling as an example of what it already excludes.

One honest caveat: We looked for a sunset or expiry on that prohibition and did not find one, and we did not retrieve the separate JADU article (§§ 66333 and following), which has its own owner-occupancy treatment. Do not read this as covering a JADU.

And there is no hearing to lose. Gov. Code § 66317(a)(1): “A permit application for an accessory dwelling unit shall be considered and approved ministerially without discretionary review or a hearing, notwithstanding Section 65901 or 65906 or any local ordinance regulating the issuance of variances or special use permits.” The word doing the work is ministerial.

A ministerial decision is made against objective standards by a person with no discretion to weigh the merits, which is why there is nothing for a neighbor to attend and nothing for a council to vote on.

Under Gov. Code § 66317(c), “No local ordinance, policy, or regulation, other than an accessory dwelling unit ordinance consistent with this article shall be the basis for the delay or denial of a building permit”.

That closing clause is the one worth carrying to the counter, because the delay most people meet is not a denial under the ADU article — it is some other ordinance, policy or internal practice being offered as the reason nothing is moving.

An existing non-conformity is not yours to cure. Under Gov. Code § 66323(a), “A local agency shall not require, as a condition for ministerial approval of a permit application for the creation of an accessory dwelling unit or a junior accessory dwelling unit, the correction of nonconforming zoning conditions.” This one is quietly enormous for older housing stock.

A house with a garage two feet from the line, a driveway that is narrower than the current standard, a rear addition permitted under a repealed rule — none of that becomes a condition of your ADU approval.

The ADU is not the moment the property is brought up to today’s zoning.

And the ADU does not drag sprinklers into the main house. Gov. Code § 66314(d)(12): “Accessory dwelling units shall not be required to provide fire sprinklers if they are not required for the primary residence. The construction of an accessory dwelling unit shall not trigger a requirement for fire sprinklers to be installed in the existing primary dwelling.” Retrofitting a sprinkler system into an existing dwelling is a five-figure job that has killed real projects.

The statute takes it off the table where the primary residence was never required to have one.

And the decision is on a clock. Under Gov. Code § 66317(a)(3), as amended by SB 543 (Stats. 2025, ch. 520, effective 1 January 2026), the permitting agency “The permitting agency shall either approve or deny the application to create or serve an accessory dwelling unit within 60 days from the date the permitting agency receives a completed application if there is an existing single-family or multifamily dwelling on the lot.” If the local agency has not approved or denied the completed application within 60 days, the application is deemed approved.

If the applicant requests a delay, the 60 days are tolled for the length of the delay.

SB 543 added a completeness clock: a written completeness determination within 15 business days, a list of what is missing, a limited resubmittal, and a deemed-complete consequence if the agency misses the determination.

An appeal of incompleteness or of a denial gets a final written determination within 60 business days (§ 66317(d)). Those are business days; the 60-day decision is not labeled as business days.

So what is left? Everything physical, and the money. That is the honest shape of an ADU feasibility question in California today, and it is almost the reverse of the one people prepare for. The rest of this page is about the half the statute does not reach.

The fee you have read about, and the bill that actually arrives

The utility connection charge is the single most common unpleasant surprise in an ADU budget, and it happens to people who did their homework — because the protection they read about is real, and it does not cover the charge they receive.

What the statute does. Under Gov. Code § 66311.5 — and note The ADU fee rules that used to sit in former § 65852.2, and then at § 66324, were renumbered to § 66311.5 by SB 543 (Stats. 2025, ch. 520, effective 1 January 2026). — Impact fees may not be imposed on an ADU of 750 square feet of interior livable space or less (or a JADU of 500 or less).

Above 750 square feet, impact fees must be proportional to those for the primary dwelling. School developer fees under Education Code § 17620 are a different statute.

Section 66311.5(c)(3) treats an ADU or JADU under 500 square feet of interior livable space as not increasing assessable space by 500 square feet.

What it does not do. Connection and capacity charges are NOT impact fees.

Under § 66311.5(b), an ADU is not a new residential use for connection-fee purposes unless it is built together with a new single-family dwelling — but the utility can still bill a connection or capacity charge, and that bill is outside the impact-fee prohibition entirely.

Why the distinction is not pedantry. An impact fee and a connection charge are levied by different bodies for different reasons under different statutes.

The impact fee is the city charging for the burden a new dwelling places on parks, roads and schools.

The connection or capacity charge is the utility charging for the physical hookup and for the share of system capacity the new unit takes.

A prohibition written against the first has no reach over the second, and reading a summary that says “no fees under 750 square feet” as covering both is how a budget acquires a five-figure hole.

The cliff on top of it. A unit at 749 square feet of interior livable space and a unit at 751 can differ by five figures of park, traffic and school charges before anyone lifts a hammer.

So there are two separate money questions hiding inside what looks like one: which side of 750 square feet of interior livable space the design lands on, and what the two utilities will charge regardless of that.

The first is yours to decide. The second is theirs to tell you, and only in writing.

What to do about it, in order. Ring the water provider and the electricity utility before you commission a single sheet of drawings.

Ask each for a written figure and for the specific question of whether the existing service is adequate or must be upsized.

If the answer is that a service upgrade is required, that number belongs at the top of the budget rather than in the contingency, and it is a legitimate reason to conclude that the honest answer for now is no.

Approval is the cheap half, and California’s own reporting proves it

A feasibility tool is really a tool for deciding where to spend the next thousand dollars. The state’s reporting says something uncomfortable about where the risk actually sits, and it is not at the counter.

The measurement. California Department of Housing and Community Development, Housing Element Annual Progress Report, Table A2 publishes Table A2 of the Housing Element Annual Progress Report.

Across 2018-2025 it carries 281,321 ADU rows from 511 jurisdictions: 178,495 permitted units against 102,356 completions, a cumulative ratio of 57%. Following the 2018–2023 permit cohort forward gives 41% with no matching completion. The method is on the research page.

Read the caveat before you use the number. These are reported APR rows, not a physical census of construction.

An unmatched permit may be a unit still under construction, an abandoned project, a unit completed without a reported certificate, or a project whose identifier changed between the permit row and the completion row.

The direction is robust; the exact rate is a reporting artefact as much as a construction one.

Why it is not simply “still building”. Among matched projects the median lag is 1 year: 29% complete in the permit year, 79% within one year and 94% within two. A project with no reported completion three years after its permit is not usually mid-build. Something stopped it.

What that does to a feasibility question. It moves it. If a large minority of people who obtained a building permit — drawings paid for, fees paid, approval in hand — did not end up with a building, then “will they let me” is the wrong risk to be spending anxiety on.

The risks worth checking first are the ones that bite after approval: the money, the utility, and whatever is physically true about the parcel.

And the money has its own gate. Putting the two together: a state can make an ADU legal to build, quickly and without a hearing, and still leave it hard to finance — because the permission is granted by a planning statute and the money is governed by a secondary-market guide that no legislature amended.

That is our analysis of the retrieved material, not a claim either enterprise makes.

Freddie Mac’s ADU page states: “ADUs on 1-, 2- and 3-unit properties must be legally permissible by jurisdiction, legal non-conforming or be located in an area without zoning.” Which is why an existing unpermitted unit is a financing problem before it is a paperwork one, and why legalizing it changes more than the file.

A 60-day statute, and medians of 162, 175 and 445 days

Both numbers are real and they measure different things. Understanding the gap is more useful than either, because most of the gap is time your own side of the table controls.

The observed distributions. Los Angeles: median 162 calendar days across 27,188 non-supplemental ADU building permits. Unincorporated Marin: median 174.5 across 472. San Francisco: median 445 across 1,713 unique issued permits.

Why this is not a finding that anyone broke the law. These are calendar days from the published application-side date to issuance, on permits that issued.

The statutory 60 days runs from a COMPLETED application, and none of the three extracts marks the completeness determination.

The gap between the two therefore includes every day an applicant spent curing an incomplete set, paying fees, answering plan-check comments, or asking for delay.

A median above 60 days is not, by itself, a finding that a city missed the shot clock.

Which is exactly what makes it useful at the feasibility stage. If most of the gap between 60 days and 162 is time spent curing an incomplete submission, answering plan-check comments and paying fees, then the schedule is substantially a function of how complete your first submission is.

The most valuable thing a designer does on an ADU may not be the design at all — it may be submitting a set complete enough that the statutory clock actually starts.

Two jurisdictions are missing, and that is a finding. City of Sacramento: The published issued-permit layer has Status_Date (the issued/status date) and Application as a permit number (e.g. RES-2600002).

No application/filed/submitted date field is published. Duration cannot be computed. City of San José: CKAN resources publish ISSUEDATE and FINALDATE. No application, filed, received or submitted date. Duration cannot be computed.

Neither publishes an application-side date, so no duration can be computed. We would rather say so than substitute a proxy from a city that does publish one.

Plan against the spread, not against the median

A median is a single number and it hides the thing you actually need for a decision: how wide the outcomes are. In all three jurisdictions the quarter that takes longest takes very much longer.

The tails. 19% of the Los Angeles permits took more than a year, and 62% of the San Francisco ones did.

At the other end, 19% of Los Angeles permits and 12% of Marin ones issued within 60 calendar days of the application-side date — so fast outcomes plainly exist, and they are not the majority.

San Francisco has an extra column worth naming. The extract also carries 952 filed applications with no issue date at all.

Those are excluded from every duration above, because a duration cannot be computed for a permit that has not issued.

That exclusion is not neutral: it removes exactly the slowest cases from the sample, which means the observed medians are, if anything, flattering.

What to do with a wide distribution. Not average it. The useful question is what your project would do if it landed in the upper quarter — whether the financing survives another two quarters of carrying cost, whether a rate lock expires, whether a family arrangement that motivated the whole thing still holds.

A feasibility answer that only works at the median is not a feasibility answer.

And it is a reason to ask about the pre-approved route. Under Gov. Code § 65852.27, A complete application for a DETACHED ADU that uses a plan preapproved in the current triennial California Building Standards Code cycle, or a plan identical to one the agency already approved in that cycle, must be approved or denied ministerially within 30 days.

By its terms that clock does not reach attached ADUs, garage conversions or JADUs unless a local program separately covers them. Whether your agency has a usable catalog is a separate question, covered on the pre-approved plans research page.

It is worth one phone call, because it is the only route in the statute that shortens the clock rather than merely enforcing it.

What this checker cannot do, specifically

Not a disclaimer. A list of the real boundaries, so you know exactly which questions you are still holding after you close the page.

It cannot see your parcel. It has no zoning map, no assessor record, no title report, no soils data, no utility service record and no overlay layer.

Every question above about easements, capacity, septic and fire zone is a question precisely because no web form can answer it, and a tool that pretended otherwise would be producing confident output from nothing.

It cannot read your ordinance. The statutory floors quoted on this page are state law.

What sits on top of them — objective design standards, local implementation, an ADU ordinance the agency has or has not updated — varies by jurisdiction, and we do not publish a 50-state table because we have not researched 50 states to that standard.

It cannot price anything. We hold no dataset of ADU bids, quotes or contracts.

Every dollar figure on this site is a named third party’s published figure, arithmetic we performed on one and have labeled as ours, or a number you typed in.

In particular it does not hold, and will not estimate, a connection or capacity charge for any utility anywhere.

That figure is the one most likely to change the decision, and inventing a magnitude for it would be worse than publishing none.

It cannot tell you whether the money will be there. That eligibility sentence is a hard gate, and it is the reason an unpermitted conversion is a financing problem rather than a paperwork problem.

It is also why Maine’s statutory amnesty for units built without municipal approval is more consequential than it first looks: legalizing the unit is what makes it financeable.

An approval and a loan are granted by different institutions under different rules, and this page touches only the first.

It cannot resolve an HOA. A recorded declaration is a private contract. States have gone opposite ways on whether covenants banning ADUs survive, and the answer for your association turns on the text of your declaration and the law of your state, read together.

And it will never say yes. Not on any combination of inputs, including all seven answered confidently. “I think yes” is a belief; the ordinance, the fee schedule and the utility are the determination. The verdict string is hard-coded and the engine test asserts it.

A national page that told you your parcel qualified would be asserting a local rule it cannot know — which is a more expensive failure than telling you honestly that you still have five calls to make.