Fire Safety · California · Analysis
Health Inspections and Fire Inspections: Two Systems, Not One
Most conversations about commercial kitchen requirements collapse two separate regulatory systems into one.
By Arthur Haggerty, Founding Editor · IKECA CECS · Published August 27, 2026
Most conversations about commercial kitchen requirements collapse two separate regulatory systems into one. It is a convenient simplification and it produces a specific, recurring blind spot.
Environmental Health
Retail food safety is governed by the state retail food law, administered by the county environmental health department — or by a municipal health department in the small number of cities that operate one independently. Subject matter is food: handling, holding temperature, sanitation, employee health and training, and facility condition as it bears on those things.
Inspection frequency is generally assigned by a risk categorization the department maintains. Results are published in most counties, in one of several formats.
Fire
The commercial cooking exhaust system, the fixed fire suppression system, detection, and egress are governed by the authority having jurisdiction for fire, applying the adopted fire code and the standards it incorporates by reference.
Which agency that is depends on the address, not the ZIP code. It may be a municipal department, an independent fire protection district, a county department, a joint powers authority, or a county department under a state cooperative agreement.
Fire findings are generally not published in a consumer-facing format anywhere.
The asymmetry
Because health results are published and fire results are not, the two obligations carry very different visibility — and therefore very different pressure.
A health finding has an immediate public consequence. It may be posted on the door. A customer can look it up. Nothing comparable exists on the fire side in most counties.
That asymmetry produces a predictable pattern: the obligation with a public consequence is resourced, and the obligation without one is deferred. Not through negligence — through the ordinary operation of attention under constraint.
The absence of a public record is not the absence of a risk. It means nobody is watching that one.
What a single number destroys
The temptation, having established that there are two systems, is to average them.
Doing so eliminates the only information that mattered. If food safety is strong and fire safety is weak, the average is comfortable — and comfortable is precisely what stops anyone from looking further. Reverse the two and the average is identical while the problem is entirely different.
One figure cannot represent two independent obligations. It can only obscure whichever is worse.
EvidLY evaluates each side against the authority that governs it, using that authority’s own methodology, and presents them separately. Where a county publishes a letter grade, you see a letter grade. Where it publishes a numeric result, you see that result calculated as the county calculates it. Where it publishes nothing, we do not invent something.
ScoreTable county pages document what each county publishes and how its methodology works, for anyone who wants to check the underlying record directly.
Sources verified by Arthur Haggerty, Founding Editor, August 27, 2026:
- Statutory basis of retail food regulation and its county administration — California Retail Food Code, Health and Safety Code §113700 et seq.
- Fire code adoption and the standards it incorporates by reference — California Fire Code, Title 24, Part 9, and NFPA 96.
- Fire authority types (municipal department, fire protection district, county department, joint powers authority, state cooperative agreement) — Fire Protection District Law of 1987 and individual authority materials.
- Independently operating city health departments (Berkeley, Long Beach, Pasadena, Vernon) — each department’s own published materials.
Full editor notes: Read the verification log.
Open study · three minutes
We’re asking California kitchens which fire and food safety records they could produce on demand. The study takes three minutes and gives you back your own snapshot — no contact details needed.
Start the study →