What EPA's Compliance Database Actually Says About Body Shops

Short answer: EPA has no inspection on record for 20,935 of the 29,646 facilities it classifies as body shops. Almost all of those still carry the status "No Violation Identified", which records that no violation has been found, not that the shop was inspected and came out clean. It is a useful record if you read it as an inspection history. It is misleading if you read it as a grade.

By Anthony Braswell for Quorum Industries LLC, The Autobody Directory · Updated 2026-08-27 · How this was written, and what the machine may not do

How this page was produced: Researched, drafted and checked with AI assistance under human direction, and signed off by the named author. How this site is written

Body shops handle solvents, paint waste, spray booth emissions and used oil. That puts many of them inside federal environmental programmes, and EPA publishes what it knows about regulated facilities through ECHO, the Enforcement and Compliance History Online system. Anyone can download the whole thing.

We did. What follows is what the file says about the industry, rather than about any one shop.

What we counted, and what the count is of

EPA distributes a bulk export called ECHO_EXPORTER.csv, 133 columns, one row per regulated site. In the snapshot we read, the file held 3,162,002 rows, of which 29,646 carry the NAICS code 811121, which is Automotive Body, Paint, and Interior Repair and Maintenance.

Two scope notes before any number below is quoted anywhere.

This is not a census of American body shops. A shop appears in ECHO only if it is in a federal or delegated environmental programme at all. Shops too small to register, or registered under a different industry code, are simply absent. The 29,646 is a count of facilities EPA has classified 811121 in this file, not a count of body shops in the United States.

These figures describe one snapshot, the ECHO Exporter as it stood on our copy dated 2 July 2026. ECHO refreshes; the numbers will move.

Seven in ten have no inspection on record

The oldest inspection date in the 811121 set is 1983. Every facility with an inspection on record has a parseable date for it, so none of the above rests on a missing field.

The label that causes the trouble

EPA gives each facility a compliance status. Across the 29,646:

Ninety nine facilities, 0.3% of the set, carry any violation status at all. None carries EPA's significant-noncompliance flag.

That looks like an industry in excellent environmental health. It is not what the data supports, because of one overlap:

20,905 facilities carry "No Violation Identified" and have no inspection on record. That is 70.5% of all 811121 facilities in the file.

"No Violation Identified" is a statement about EPA's records. It says no violation has been identified. Where there has been no inspection, that is a description of the absence of an inspection, and it should not be read as a finding of compliance. The 99 facilities with a violation status are, definitionally, drawn almost entirely from the 8,711 that someone actually visited.

The honest way to read any single shop's row is as two facts held together: when was it last inspected, and what was found. Either one alone will mislead you.

Inspection coverage varies enormously between states, and this is a fact about the database

Among the fifteen states with the most 811121 facilities in this file:

California holds the largest 811121 population in the file by a wide margin and the lowest recorded inspection rate of these fifteen.

We are not saying California inspects fewer body shops than Pennsylvania. That is the inference the table invites and it is not one this file can support. ECHO aggregates federal programmes and state programmes that report into federal systems, and what a state sends to EPA varies by programme, by agency and by era. A low rate here is consistent with less inspection, and equally consistent with inspections that happen under a state programme which does not flow into ECHO. What the table establishes is how much inspection history ECHO holds per state. What causes the variation is a separate question this file does not answer.

What these shops are actually regulated under, and it is mostly waste

A body shop can appear in ECHO under several federal programmes, and the file records which. Every one of the 29,646 is in at least one:

The combinations are the more useful cut: 20,744 are in RCRA and nothing else, 4,902 are in the Clean Air Act and nothing else, 3,621 are in both, and 193 add the Clean Water Act to RCRA.

So when this page says "EPA's record", it is mostly a hazardous waste record. Four in five of these facilities are federally visible because of spent solvents, contaminated thinners and sanding dust, not because of what comes out of the booth. That is worth knowing before you read a row: the inspection that did or did not happen was, for most of these addresses, a waste inspection.

One thing an air flag is not: a determination under the paint stripping and surface coating rule that governs booths. That rule is not in this file at all. It is a separate question with a separate answer.

Fewer than one in twenty has been inspected in the last five years

The file carries two different kinds of inspection field and they are easy to confuse, so it is worth being exact. EPA's own column dictionary settles it. The counts cover a window:

"The number of inspections/compliance evaluations that have occurred at the facility, under the CAA, CWA, and RCRA, within the last five years."

The date does not:

"The date on which most recent inspection of the facility took place."

So the 8,711 figure above is all-time, and a separate, smaller number answers a sharper question. 1,319 facilities, 4.4% of the 29,646, carry an inspection inside that five-year window. Fewer than one in twenty.

We checked EPA's definition against the data rather than taking it on trust. Of the 1,319, all but one had their last inspection inside five years of the snapshot, a 99.9% fit. Testing a three-year window instead gives 69.7%, which is the wrong shape. And no facility carries a count without also carrying a date, so the two fields are consistent rather than contradictory.

This also disposes of a theory worth recording because it was wrong. We first guessed the counts were federal-led work while the dates included state-led work. That is not it: 1,204 of the 1,319 have a state inspection date only. The difference was never federal versus state. It was five years versus all time.

Two things we checked because a reader would ask

Is the never-inspected share just closed shops? ECHO keeps historic facilities, so a fair objection is that the 20,935 are largely sites that no longer exist and were never inspected because there was nothing to inspect. The file carries an active flag, so this is measurable rather than arguable. 24,195 of the 29,646 are flagged active. Among those, 17,130 have no inspection on record, which is 70.8%, against 70.6% across the whole set. Restricting to facilities EPA still flags as active does not move the finding.

Who does the inspecting? ECHO records the last EPA inspection date and the last state inspection date in separate columns, and the split is lopsided:

Those three sum to the 8,711 with any inspection on record, which is the same total the first pass produced from a different column. 93% of the inspection history in this set carries a state date.

That matters for the table above. ECHO's 811121 inspection record is overwhelmingly a record of state activity reported into federal systems, not of EPA visits. So a state's row reflects the combination of what that state does and what it reports, and the two cannot be separated from this file. California's 148 break the same way: 138 state-only, 6 EPA-only, 4 both.

Almost none of them are large hazardous waste generators

If four in five of these facilities are in ECHO because of hazardous waste, the next question is how much waste. RCRA sorts generators into tiers, and the tier decides how much federal obligation a shop carries. That field is not in the ECHO export, but it is in EPA's RCRAInfo download, so we went and got it.

A different population, and it matters. These are the 12,610 handlers RCRAInfo itself classifies as NAICS 811121, not the 29,646 facilities in the ECHO set above. The two are built from different NAICS sources and should not be quoted against each other.

88.4% are very small quantity generators or not generators at all. Only 11.3% sit in the tiers that carry meaningful ongoing federal duties, and 36 body shops in the entire country are large quantity generators.

That is most of the answer to the question this page opened with. A regulatory population that is seven-eighths very-small or non-generating is not one you would expect to see inspected often. The thin inspection record is substantially a description of how the rules are scoped, not evidence that anyone is asleep.

EPA's code list runs 1 for large, 2 for small, 3 for conditionally exempt small, N for not a generator. "Conditionally exempt small quantity generator" is the older name; the 2016 Generator Improvements Rule renamed that tier very small quantity generator, and we use the current name above. We did not take the mapping on trust: the file carries a second, independent column naming the universes, and it matches the codes exactly in all three tiers.

Penalties are rare and concentrated

58 facilities, 0.2% of the set, carry any penalty total. Between them they account for $16,882,972. The median penalty among those 58 is $6,750; the largest single facility total is $5,300,000.

So the money is real but it lands on very few addresses, and one facility accounts for roughly a third of the industry total in this file. An industry-wide average penalty would be a meaningless number and we have not calculated one.

How to use this if you run a shop

Your ECHO row is public and permanent, and it is one of the few environmental records a customer, an insurer or a buyer can look up without asking you. Three things follow.

Look yourself up. Facilities are searchable by name and address at echo.epa.gov, and the row will show your last inspection date, inspection count and status.

If your row says "No Violation Identified" with no inspection date, understand what you are looking at. It is not a clean bill of health you can point to, because it records that nobody has been. Anyone reading it carefully will see the same thing.

If the row is wrong, the facility data comes from the programme that reported it, so corrections go through that programme rather than through ECHO itself.

How to use this if you are choosing a shop

An ECHO record is worth checking and worth putting in proportion. A shop with a recent inspection and no violation recorded against it has at least been examined by a regulator without a violation being identified, which is genuinely informative. A shop with no inspection record has told you nothing either way, and the majority of shops in this file are in that position.

What an ECHO row is not is a repair quality signal. It concerns hazardous waste, air permits and water discharge. A shop can be immaculate on all three and still fit an aftermarket part your manufacturer prohibits, or skip the calibration your car needs after a windscreen replacement.

Sources and method, and the limits of both

Two sources, both United States Government works in the public domain.

EPA's ECHO Exporter, echo.epa.gov/files/echodownloads/echo_exporter.zip, member ECHO_EXPORTER.csv, 133 columns, for everything about the 29,646.

EPA's RCRAInfo download, echo.epa.gov/files/echodownloads/rcra_downloads.zip, for the generator tiers: RCRA_NAICS.csv filtered to NAICS 811121 and joined to RCRA_FACILITIES.csv on facility ID and activity location, giving 12,610 handlers. Column definitions come from EPA's own ECHO Exporter column dictionary rather than from inference.

Selection was an exact token match on the FAC_NAICS_CODES field, which holds space-separated codes; a facility counts if 811121 appears as its own token, so a longer code containing those digits does not match. Column positions were resolved from the file's own header rather than assumed, because a guessed column yields a precise and wrong answer rather than an error. Every figure above was computed from the file in one pass and reconciled: the status counts sum to the facility total, and inspected plus never-inspected equals the facility total.

Five limits worth stating plainly.

The snapshot is dated 2 July 2026 and ECHO changes.

An inspection recorded in ECHO is an inspection some programme reported to EPA. Absence of a record is not proof that no inspection occurred.

NAICS classification in ECHO is assigned per facility and can be wrong or stale, so a shop may sit under a different code and be missing from this set entirely.

The generator tiers describe a different population, the 12,610 handlers RCRAInfo classifies 811121, and are not a breakdown of the 29,646.

Finally, we did not attempt to match these facilities to individual businesses for this piece. The figures are about the populations in EPA's files, not about any named shop.

Corrections

2026-08-27. An open question on this page is now answered, and one of our guesses about it was wrong. An earlier version said we would not publish a per-programme inspection count because it did not reconcile: 8,711 facilities carried an inspection date while only 1,319 carried a count. It said we had tested the theory that counts were federal-led and dates included state-led work, and that the theory was wrong. It was. The actual answer, from EPA's own column dictionary, is that the count columns cover the last five years while the date columns are all-time. Confirmed against the data: 1,318 of the 1,319 fall inside a five-year window, and a three-year window fits only 69.7%. The section now reports the five-year figure instead of withholding it.

2026-08-27. The opening sentence and two headings treated an absence of record as a fact about the world. They read "EPA's public database lists 29,646 body shops and has never inspected 20,935 of them", and the table and a section heading said "never inspected". No figure has changed. The wording has, because this page's own limits section says an inspection recorded in ECHO is one that some programme reported to EPA, and absence of a record is not proof that no inspection occurred. Saying EPA "has never inspected" a shop asserts something the file cannot establish, which is the exact misreading this page was written to warn against. It now says "no inspection on record" throughout. The same sentence also called 29,646 ECHO facilities "body shops" while the section below correctly calls them facilities EPA has classified 811121, and that is fixed too.

2026-08-27. "Passed" was too strong. The guidance for choosing a shop said a recent inspection with no violations meant a shop "has been looked at by a regulator and passed". An inspection with no violation identified is not a certification of overall compliance, and the sentence now says so in longer and duller words.

If you find an error in the figures above, the underlying file is public and we will publish the correction here with the date and the wording it replaced.

General consumer information: not legal, insurance, or financial advice. Requirements, coverage, and practices vary by state, policy, and manufacturer.

Where this fits

Each link says what it is for. We add one only when a reader on this page has a real reason to need that page next.

• The 6H Rule at 18: Painter Certification, 98% Filters, and the Part That Went Electronic (the rule that actually governs your spray booth and paint stripping, which an enforcement record does not tell you whether you are meeting)
• Aerosol Cans Became Universal Waste in 2020, and in Fourteen States They Still Are Not (the generator category on this page is set by what you count, and in most states aerosol cans no longer have to be counted)