What is the US dental claim denial rate? The only one that matters is yours.
We went looking for a national dental claim denial rate and could not find one published by any primary source: not the American Dental Association, not the Centers for Medicare and Medicaid Services, and not the peer reviewed literature. The figure usually quoted for dentistry comes from an analysis of medical insurance.
Which is less of a problem than it sounds, because a national average was never going to tell you which payer is declining your crowns. Your practice already produced the only rate that can be acted on: it is in the claims sitting in your practice management software right now. The rest of this page is where the quoted number comes from, why three different events keep getting counted as one, and how to read your own.
Where the number people quote comes from
Trace the denial rates in circulation for dentistry and they run out quickly. Some are published by companies that sell dental billing services, with no organization, study, sample or year named anywhere on the page. Others are real, carefully produced figures, measured on something other than dental insurance and relabeled on the way down. The most widely repeated one is the second kind.
The claim denial rate most often quoted for dentistry comes from an analysis of medical marketplace plans, and that analysis says in its own methods that it excludes standalone dental plans.
It is a real figure, and it is about medical coverage. It was never a measurement of dental claims.
Our readingWe went looking for a national dental denial rate and could not find one published by anyone: not the ADA, not CMS, not the peer reviewed literature. The only denial rate that describes a practice is the one in its own claims.
KFF, methodology for its analysis of marketplace claim denials, 2026Could a real one be calculated?
For one slice of the market, yes. The federal transparency file that the medical analysis draws on does carry claims received and claims denied counts for standalone dental plans, so a denial rate for marketplace dental coverage is derivable from primary data by anyone willing to do the arithmetic.
That is not the same as a national dental denial rate, and it is not what anyone is quoting. Marketplace standalone dental plans are a small and unrepresentative corner of how Americans get dental coverage, so a figure computed there would describe that corner and nothing else. What does not exist is anyone having published a rate for dentistry as a whole.
Three different things, counted as one
Part of why the figures in circulation disagree with each other so badly is that they are not all counting the same event. Most reporting pours these three into one bucket, and they have different causes and different fixes.
The claim never reached adjudication. A clearinghouse or the payer's front door bounced it on a format or eligibility problem, often within hours. Nothing was decided about the treatment.
The payer has the claim and wants something more before it decides: a narrative, a radiograph, a perio chart. It is pending, not refused, but it sits in the same pile on most reports.
The payer adjudicated the claim and declined to pay some or all of it. This is the only one of the three that is actually a denial, and it is the one a rate should count.
So what should a practice do instead?
Measure your own, and then work it. The counts are already in your practice software, and once you have them the useful questions open up immediately: which payer, which procedure, which reason, and which of those are yours to fix rather than the plan's. That is a list your team can work on a Tuesday. A national average is not.
Or have it done for you
Practice Evolved connects to the practice management software your office already runs, reads it without ever writing to it, and keeps this count for you: what was billed, what came back denied, which payer and which reason it came back on, and which of them are still inside their appeal window. The rate is the easy part. What the dashboard is actually for is the list underneath it.
We wrote this page because we went looking for a benchmark to put on that dashboard, could not find one that survived checking, and decided to publish that rather than repeat it. Every number your own dashboard shows you comes out of your own database, so you never have to take ours on trust either.