Can Car Insurance Deny a Claim?

The Clock Running While You Wait

Silence feels like the company doing nothing. Under the model regulation, silence is the company burning through deadlines. Each interval below is an obligation the insurer owes you, and each one is a specific, checkable date you can put in an email.

One carve-out matters. Where there is a reasonable basis, supported by specific information available for review by the insurance regulator, to suspect that the first-party claimant fraudulently caused or contributed to the loss, the insurer is relieved of those fixed intervals — but it still must advise the claimant of acceptance or denial within a reasonable time for a full investigation after receiving a properly executed proof of loss.^[1] That is the provision behind a claim that suddenly goes quiet for months. It does not erase the duty to answer; it converts a hard number into a reasonableness test.

If your claim is sitting open well past these marks, our companion research on how long a car insurance claim stays open walks through the full lifecycle from first notice of loss to final settlement.

Denial Ground One: The Loss Was Never Covered

Start with the covered case, because the exclusions only make sense against it. A collision coverage promises to pay for direct and accidental physical damage — a sudden external event that happens to an otherwise sound vehicle. The energy comes from outside the car, arrives in an instant, and leaves damage traceable to that instant. Everything in the exclusion list is an attempt to describe damage that fails one of those three tests.

Gradual deterioration fails the “sudden” test. A control arm that has absorbed forty thousand cycles of road load does not fracture because of one pothole; it fractures because the crack that finally ran through it had been advancing a fraction of a millimeter at a time. Forensic engineers read that history off the fracture surface — smooth, polished bands and concentric beach marks mean fatigue, while a single jagged, deformed tear means one high-energy impact. When the analysis lands on fatigue, the mechanical-breakdown and wear-and-tear exclusion applies and the claim is denied. Our separate report on whether car insurance covers wear and tear works through that boundary, including the ensuing-loss rule that still pays for the crash a failed part causes even when it will not pay to replace the part.

Intentional damage fails the “accidental” test, and commercial use fails a different test entirely: it is not the kind of risk the premium was priced for. The Texas Department of Insurance puts it in its consumer auto guide under the heading “Most policies don’t cover” — accidents that happen while you are driving for a ride-hailing service or delivering food or other items for a fee, along with business driving, racing, and damage you caused intentionally — while noting that coverages vary by policy.^[3] Read that qualifier literally, because it is where the fight happens. Whether a driver logged into a delivery app at the moment of impact is covered turns on the exclusion as written in that policy, on what the driver was actually doing when the crash occurred — cruising unassigned, en route to a pickup, or carrying the order — and on whether a rideshare or delivery endorsement is attached. The default on an unendorsed personal policy is that the loss is excluded; the endorsement is what changes the answer. Our research on car insurance requirements for DoorDash drivers covers the three-period framework that closes that gap.

The driver can also be the excluded item rather than the activity. The same Texas guide notes that while most policies cover you, your family, and people driving with your permission, you should read the policy to learn who is covered and whether anyone is excluded from coverage.^[3] That last clause points at the named-driver exclusion endorsement — a signed document removing one specific person from the policy, usually a household member whose record would otherwise price the policy out of reach. Texas draws a fine line here that is easy to misread: House Bill 259 barred insurers from issuing or renewing “named driver policies” — policies that do not cover a household resident unless that person is named — on or after January 1, 2020.^[4] An excluded-driver policy is the functional opposite and was not swept up in that ban: as the department explained when it wrote the implementing rule, an excluded driver policy generally covers all drivers except those specifically excluded by name, and excluding specific drivers by name does not by itself make a policy a named driver policy.^[17] So the narrow, by-name exclusion survived. If an excluded driver takes the wheel and crashes, the denial follows from the endorsement the policyholder signed.

Denial Ground Two: The Loss Was Covered, But You Broke a Condition

This is the category that surprises people, because the accident itself is plainly covered. A policy does not only describe what it pays for; it describes what you must do after a loss so the company can verify it. Those duties are conditions, and failing one can forfeit an otherwise valid claim.

The New Jersey Department of Banking and Insurance lists the core set for a driver filing an auto damage claim with their own insurer: immediately report all losses, supply a police report for a theft or hit-and-run, make the damaged vehicle available for inspection by the insurance company before you have it repaired, and protect your vehicle from further damage.^[5] Each one exists for a mechanical reason. Repairing before inspection destroys the evidence the adjuster needs to separate crash damage from pre-existing damage. Leaving a shattered window open to a rainstorm lets a covered glass loss grow into an uncovered water-damage loss — the company pays for the glass and declines the interior, because the second loss was not the crash, it was the days after it.

Storage is the version of this that costs real money quietly. New Jersey’s guidance sets a specific procedural protection: your insurance company must give you three working days’ notice before it stops paying storage charges, so you have time to move the vehicle and avoid continuing fees.^[5] Read the flip side of that sentence. Once the notice runs, every additional day the car sits at the tow yard is yours. A partial denial of storage fees is not the insurer being difficult; it is the end of a clock you were told about in writing.

The heaviest condition is the duty to cooperate, and where a policy contains one, its sharpest expression is the examination under oath — a formal, recorded proceeding in which the insurer’s counsel questions the policyholder on the record and under oath, before a court reporter. The legal weight behind it is old. In Claflin v. Commonwealth Insurance Company, decided in 1884, the Supreme Court treated a policyholder’s knowingly false statements under a required examination as fraud and a breach of that policy’s conditions, working a forfeiture of the claim even where the insured said the motive was self-protection rather than deceiving the insurer.^[6] The modern consequence runs along the same line without being automatic: a material, willful refusal to sit for the examination, or a knowingly false answer in one, gives the insurer a coverage defense that never requires it to argue about the crash at all. How far that defense reaches — whether the insurer must also show prejudice, and what counts as material — is a question of the policy language and of your state’s law, not a single national rule.

Denial Ground Three: The Policy Itself Is Unwound

The most severe outcome is not a denied claim. It is a policy treated as though it never existed. When an insurer concludes that an application contained a material misrepresentation — the wrong garaging address, an undisclosed household driver, a hidden commercial use — it may seek rescission: the contract is voided from inception, the premiums are returned, and every claim under it falls with the contract.

Courts have been unwilling to let that remedy run unchecked when someone else was injured, because the injured party had no part in the lie. California drew the line in 1969. In Barrera v. State Farm Mutual Automobile Insurance Company, the California Supreme Court held that an auto liability insurer must conduct a reasonable investigation of an applicant’s insurability within a reasonable time after issuing the policy, and that an insurer which fails that duty cannot then rescind as a defense against an injured third party who obtained a judgment against the insured driver.^[7] The reasoning is a cost-allocation argument: an insurer that can collect premiums indefinitely while postponing verification has no incentive to verify at all.

Michigan reached a different structural answer in 2018. In Bazzi v. Sentinel Insurance Company, the Michigan Supreme Court confirmed that its earlier decision in Titan Insurance Co. v. Hyten had abrogated the judicially created innocent-third-party rule, so an insurer is not categorically barred from rescinding a policy obtained by fraud even where an innocent third party has a claim. But the court did not make rescission automatic either: because rescission is an equitable remedy, a trial court considering it as against an innocent third party must balance the equities between the insurer and that third party.^[8] Two states, two doctrines, one practical lesson — whether a misrepresentation wipes out coverage for a bystander is a question of state law, not of the policy language.

The Evidence Behind a Modern Denial

An adjuster in 1995 had a police report, photographs, and two conflicting statements. An adjuster today can have the car’s own recording of the seconds before impact. Most passenger vehicles carry an event data recorder — a microprocessor, typically housed in the airbag control module, that continuously buffers sensor readings and permanently writes them out when a deceleration threshold suggests a crash.

Federal regulation does not require a vehicle to have one, but for the vehicles it covers it dictates precisely what an installed recorder must capture. Part 563 reaches light vehicles — passenger cars, multipurpose passenger vehicles, trucks, and buses with a GVWR of 3,855 kg or less and an unloaded vehicle weight of 2,495 kg or less — manufactured on or after September 1, 2012, and only if they are equipped with an event data recorder; walk-in van-type trucks and vehicles built exclusively for the U.S. Postal Service are carved out.^[9] Inside that class, Table I of 49 CFR § 563.7 fixes the data elements, the recording interval, and the sample rate — and the numbers are unforgiving of a story that does not match.

Read what that table actually does to a disputed claim. Longitudinal delta-V is defined in the regulation as the cumulative change in the vehicle’s velocity along its longitudinal axis, recorded every 0.01 seconds from crash time zero.^[9] It is a velocity measurement, not an energy measurement — kinetic energy depends on the vehicle’s mass as well as its speed change — but it is the single number crash reconstructionists lean on hardest, because for a given vehicle it scales with how hard the impact actually was. A claimed high-speed rear-end collision that produced a small recorded delta-V is a claim whose physics do not support its damages. Throttle position and service-brake status, sampled twice a second across the five seconds before impact, answer whether the driver ever tried to stop. Seat-belt status at one second before the crash answers a question no witness statement resolves.

That data is not the insurer’s to take. The Driver Privacy Act of 2015, enacted as part of the FAST Act, provides that data retained by an event data recorder is the property of the vehicle’s owner — or the lessee, for a leased vehicle — and may be retrieved only under enumerated exceptions: a court or administrative authorization, the owner’s or lessee’s written, electronic, or recorded audio consent, an authorized federal safety investigation or inspection, retrieval to determine the need for or to facilitate emergency medical response after a crash, or traffic-safety research with identifying information protected.^[10] In practice the second exception is the one that matters, and it usually arrives as a consent form inside a stack of routine claim paperwork.

When “Denied” Is Really a Fight About the Number

A large share of what drivers describe as a denial is not a coverage refusal at all. The company agrees the loss is covered and disagrees about what it is worth — most often because it declared the vehicle a total loss and wrote a check for actual cash value instead of paying to repair. New Jersey’s regulator states the trigger: a vehicle is a total loss when the cost to repair equals or exceeds the vehicle’s actual cash value on the date of loss, and companies often make that call at around 80% of actual cash value because hidden damage tends to surface once a shop opens the car up.^[5] How that value is built is covered in our work on how insurance determines the value of a totaled car and on how adjusters determine a car’s value.

When the disagreement is about dollars, the policy already contains the tool. The appraisal clause lets either side demand an appraisal: you choose and pay for an appraiser, the company chooses and pays for its own, the two select a neutral umpire, and a decision by any two of the three binds the parties.^[5] It resolves a valuation dispute without pleadings, subpoenas, or a hearing.

It will not resolve a coverage denial, and confusing the two wastes months. The Texas Supreme Court framed the boundary in State Farm Lloyds v. Johnson, decided July 3, 2009: the scope of appraisal is damages, not liability — appraisal clauses contractually resolve the amount of the covered loss, and an appraisal award does not by itself show that an insurer breached the policy by failing to pay.^[11] So the test is simple. If the letter says the loss is covered but worth less than you think, invoke appraisal. If the letter names an exclusion or a breached condition, appraisal is the wrong instrument and the next step is regulatory or legal.

What Turns a Wrong Denial Into a Liability

A denial the insurer got wrong is a breach-of-contract problem. A denial the insurer had no reasonable basis to issue is something heavier — and states reach that second category by two different routes. Florida and Pennsylvania wrote a private statutory remedy directly into their codes. California did not: its unfair-practices statute is enforced by the Insurance Commissioner, and the policyholder’s own lawsuit comes from the common law instead.

Note the structural differences, because they change what you do first. Florida builds a cure period into the statute: the claimant files written notice with the department and the insurer, and no action lies if within sixty days the insurer pays the damages or corrects the circumstances giving rise to the violation.^[13] That notice is not a formality to skip; it is the event that starts the clock and, frequently, the event that gets the claim paid.

Pennsylvania attaches the penalty to the money rather than to a cure. On a finding of bad faith, the court may award interest on the claim from the date it was made at the prime rate plus 3%, award punitive damages, and assess court costs and attorney fees against the insurer.^[14] California, by contrast, enumerates the conduct itself — sixteen unfair claims settlement practices, prohibited when knowingly committed or performed with such frequency as to indicate a general business practice, including the failure to promptly provide a reasonable explanation of the policy basis for a denial.^[12] What that statute does not do is hand you a lawsuit. In Moradi-Shalal v. Fireman’s Fund Insurance Companies the California Supreme Court overruled Royal Globe and held that the Legislature “has not manifested an intent to create such a private cause of action” under section 790.03(h), leaving enforcement of the statute to the Insurance Commissioner. In the same passage the court kept the policyholder’s own remedy open: courts retain jurisdiction over common law actions on traditional theories including, as to the insured, breach of contract or breach of the implied covenant of good faith and fair dealing.^[18] So the practical difference is where the claim comes from, not whether one exists: a Florida or Pennsylvania policyholder sues under a statute, a California policyholder sues on the covenant implied in every insurance contract, and the § 790.03 violation is the complaint you file with the Department of Insurance.

How Regulators See What You Cannot

No state insurance department reads every denial letter. They read patterns instead. The NAIC’s Market Conduct Annual Statement (MCAS), created in 2002 with eight states and two lines of business and moved to a centralized collection system in 2011, is a uniform system for collecting market-related claims and underwriting data so states can monitor how companies actually behave. As of the 2024 data year it covers thirteen lines of business, private passenger auto among them, and regulators use it to monitor, benchmark, analyze, and regulate insurer market conduct.^[15]

The private passenger auto data call makes the denial pattern legible. Insurers report the number of claims closed during the period with payment and the number closed without payment, broken into aging buckets of 0–30, 31–60, 61–90, 91–180, 181–365, and beyond 365 days; a separate count of claims closed without payment because the amount claimed fell below the deductible; the median days to final payment; and the number of lawsuits closed with consideration for the consumer — defined as a lawsuit in which a court order, jury verdict, or settlement delivered more to the claimant than the insurer had offered before suit was brought.^[16]

That last element is the tell a single policyholder can never see. A company whose pre-suit offers routinely get beaten in court is a company whose denial and valuation decisions are systematically low — and that is the kind of signal that triggers a targeted market conduct examination rather than a form letter.

What to Do With a Denial Letter

The sequence below follows the structure of the rules above rather than a generic escalation ladder. Each step either satisfies a condition the insurer is entitled to enforce or forces the insurer to commit to a position in writing.

1. Get the denial in writing, with the provision named. The model regulation requires a written denial that references the specific policy provision, condition, or exclusion relied on.^[1] If the letter you received is generic, request the provision by name and in writing.
2. Read that provision against your actual facts. Exclusions are narrowly worded. A wear-and-tear exclusion addressed to damage due and confined to a failed part does not reach the collision that failure caused.
3. Supply whatever the insurer says is missing.If the denial rests on a post-loss duty — an inspection, a police report, documents, an examination under oath — curing the breach is faster and cheaper than litigating it, and it removes the company’s defense entirely.^[5]
4. Separate a valuation dispute from a coverage dispute. If coverage is conceded and only the amount is contested, demand appraisal. If coverage itself is refused, appraisal cannot reach it.^[11]
5. File a complaint with your state insurance department. Under the model regulation the insurer owes the department an adequate written response within twenty-one days of its inquiry, and your file feeds the market conduct data regulators use to spot patterns.^{[1] [15]}
6. Check your state’s bad-faith procedure before suing. Florida requires a written civil remedy notice and a sixty-day cure window before an action lies; other states impose different prerequisites and different proof standards, and some — California among them — route the policyholder to a common-law bad-faith claim rather than a statutory one.^{[13] [18]}