Search this question and every result points at the same free tool. None of them quote what that tool says about itself.
The NICB’s own page, in italics, under its own search box:
“While VINCheck is a useful free tool, it is not a comprehensive vehicle history report and should not be relied upon when purchasing a vehicle.”
That sentence is the honest starting point. Here is what each available check actually covers, where the holes are, and what a dealer buying ten cars a month should do instead of what the listicles say.
The scale of the problem, from the federal number
NHTSA publishes the figure. “In 2025, more than 650,000 vehicles were stolen in the United States.” Put another way, on the same page: “A motor vehicle was stolen every 48 seconds in the United States in 2025.” It adds that summers prove to be the worst season for vehicle theft.
Worth noting what that page does not contain. NHTSA gives advice to theft victims and describes anti-theft devices, and offers no guidance at all to someone deciding whether the car in front of them is stolen. Its only VIN-adjacent rule is about manufacturers: regulations require them to label major vehicle component and replacement parts so authorities can better attempt to trace and recover stolen parts.
So the buyer is on their own, with three free sources and a paid one.
Check one: NICB VINCheck, and exactly what it sees
Start with what it is for. VINCheck lets you enter a VIN to instantly learn whether the vehicle has been “reported as stolen but not recovered or reported as a salvage or flood-damaged vehicle by participating NICB member insurance companies.”
Every qualifier in that sentence is load-bearing.
It is insurance data, not police data. “VINCheck cross-references the VIN entered against participating insurer’s theft and salvage records.” And then, explicitly: “VINCheck does not query law enforcement records or records of insurance companies that elect not to participate in VINCheck.”
It covers five years. The database holds vehicles “reported as stolen within the past five years and remain unrecovered or declared as salvage.” A car stolen six years ago and never recovered is outside the window.
It covers most, not all, of the industry. The database “represents 92.49% of the earned insurance premium in the U.S.”, a figure the page dates to August 2026. Which leaves the rest.
The NICB does not vouch for it. From the terms of use: “The NICB has not independently verified the information in the database and cannot, therefore, vouch for the accuracy of this information.” Followed by three more admissions in the same list: “There are many other vehicles that have been reported as stolen which are not included in this database.” A vehicle may have been “incorrectly reported as having been stolen, or was stolen and recovered, but has not [been] removed from this database.” And: “It is also possible that there are vehicles with duplicate VINs and/or VINs incorrectly entered.”
A clean result proves nothing on its own. The NICB’s own FAQ answer to whether a clean result guarantees the car is safe to buy: “A clean result does not guarantee that the vehicle you are searching for has a clean history.”
One mechanical detail that quietly produces false comfort: “You need to input a correct 17-digit VIN to potentially match against any insurance theft, salvage, or flood records. If the VIN is incorrect, a match record may not be identified.” A typo returns the same reassuring nothing as a clean car.
The five-a-day problem nobody writes about
Here is the line that decides whether this tool is usable in a business.
“You can conduct a maximum of five searches within a 24-hour period per IP address.”
Per IP address, not per person. So a lot with an owner, a buyer and a bookkeeper all working off the same office connection does not get fifteen checks. It gets five, shared.
Five is fine for a consumer buying one car. It is nothing for a dealer walking an auction lane with forty cars on the run list. That single limit is why the free tool cannot be the process, and why “just run it through NICB” is advice written for someone else’s situation.
Check two: NMVTIS, and what it actually reports
The federal system exists partly for this. Among its stated purposes: “Prevent the introduction or reintroduction of stolen motor vehicles into interstate commerce.”
But read what a consumer report contains. NMVTIS says the report is “intentionally concise” and “intended to provide data on only five key indicators associated with preventing fraud and theft.” The four results it lists are title and brand history, the latest reported odometer readings, insurer salvage or total loss determinations, and transfers to an auto recycler, junk yard or salvage yard.
Theft is not among them. The page’s own wording for theft is a hedge: consumers can find title, odometer and brand history “and, in some cases, historical theft data.”
Its coverage has the familiar gap: 87 percent of the U.S. vehicle population is in NMVTIS, more than half the states report data into it, and consumers “should be aware of the possibility that a search may yield a false negative indication.” The same gaps, it adds, can exist in private vehicle history databases.
There is one theft-adjacent quirk worth knowing before you misread a report. A stolen vehicle not recovered within a 30-day period may be labelled total loss by the insurer, and if it is later recovered, “the ‘total loss’ determination is not removed.” So a total loss brand on an undamaged car is sometimes a theft recovery, not a wreck. What is a salvage title covers how that status then behaves.
A detail that matters if you hold a dealer licence: NMVTIS states that consumers cannot receive NMVTIS vehicle history reports from Carfax, DMVDesk or Experian, because those entities provide information only to car dealerships. The channel you can use as a licensed dealer is not the one your retail buyer can use. Carfax vs AutoCheck vs NMVTIS sets out which check answers which question.
Check three: your state may publish a free search
This is the one most guides miss entirely, and it is the only one of the three that draws on law enforcement records rather than insurance records.
Florida runs a public access system on the Florida Crime Information Center, with searchable categories for stolen vehicles, stolen vehicle parts, stolen licence plates and stolen decals. Its stated purpose: “This site is designed to provide the public with the opportunity to assist law enforcement agencies in locating stolen property.”
Read its limits before you lean on it. “This site contains records from Florida law enforcement agencies that have been selected for public access by those agencies.” And: “Other records may exist that have not been selected for public access by those agencies.” It is refreshed from the FCIC files every 24 hours, and it holds Florida records only.
Not every state publishes one, and the ones that do differ in what they expose. It is worth ten minutes to find out whether yours does, because it is free, it is a different data source from the other two, and a hit there is worth more than a clean result anywhere.
The one that takes the car and the money: cloning
If you only read one section, read this one. It is also the fastest-growing search in this whole topic.
The NICB sets out the method in five steps, in its own words:
- “Copy a VIN from a legally owned vehicle parked on the street, in a lot, or at a dealership”
- “Create a counterfeit VIN tag and corresponding fake ownership documents”
- “Steal a vehicle like the one with the legitimate VIN”
- “Replace the stolen vehicle’s VIN with the counterfeit tag, making it appear identical to the legitimate vehicle”
- “Sell the ‘cloned’ vehicle to an unsuspecting buyer”
Notice what this defeats. Every check above runs on the VIN. A cloned car carries a VIN that belongs to a real, legally owned, never stolen vehicle. It comes back clean because the number is clean. The federal guidance confirms the effect directly, describing an operation in which roughly 250 cloned stolen vehicles were recovered: because the paperwork showed legitimate VINs, “law enforcement could not detect the vehicles’ true stolen status.” A separate Florida case involved a ring cloning more than 250 cars valued at $8 million.
Why it keeps working, according to the NICB: “Because most licensing agencies don’t check for duplicate ownership when someone surrenders an out-of-state title, thieves rarely get caught.”
And what it costs the person holding the keys:
“Unfortunately, the vehicle is never legally yours, even if you paid for it. If the VIN switch is discovered, the car will be confiscated and returned to its rightful owner or to the insurance company that paid a theft claim on it.”
Not a discount. Not a hard conversation with a customer. The whole car and the whole purchase price, gone, after you have already paid for transport and reconditioning. That is the only risk in used-car buying that can take 100% of a deal.
What the NICB tells you to do about cloning
Four steps, verbatim from its protection list, and three of them are things most buyers skip:
- “Check the vehicle’s VIN with appropriate government agencies or your state bureau of motor vehicles”
- “Conduct a title search of the vehicle”
- “Analyze the ownership pattern for any late-model vehicle with no lien holder”
- “If possible, ask your insurance company to inspect the vehicle prior to purchase”
The third one is the sharpest and the least obvious. A nearly new car owned outright, with no lender ever recorded against it, is unusual. Most late-model cars are financed, and a lienholder in the history is a bank that checked the paperwork before handing over money. Its absence is not proof of anything. It is a reason to look harder.
What a rebuilt title examination already catches
If the car you are looking at went through a state rebuild examination, part of this work has been done for you, and it is worth knowing exactly which part.
New York says it plainly: “The required examination is part of the NY State Auto Theft Prevention Program,” and state law requires the DMV to determine whether a rebuilt salvage vehicle is stolen or includes stolen parts. At the appointment the owner must describe the repair work and the parts that were replaced and provide the original receipts for the replacement parts. Virginia frames its own examination the same way, as confirming that the vehicle and its parts are in safe operating condition and have not been stolen.
So a rebuilt title is, among other things, evidence that a state investigator looked at this specific car for stolen parts. That is a real check, and it is one more reason the brand is not simply bad news. Should you buy a car with a rebuilt title works through the rest of that decision.
If something comes back a hit
Both authorities say the same thing, and both say it emphatically, because this is where buyers make the situation worse.
The NICB: if a vehicle is reported in the database as stolen but not recovered, “you should verify with the NICB or your local law enforcement agency that a stolen vehicle report is active before taking any action regarding the vehicle.”
Florida’s FDLE, in capitals on its own home page: “IF YOU HAVE INFORMATION CONCERNING STOLEN PROPERTY, WANTED PERSONS OR MISSING PERSONS, DO NOT TAKE ACTION ON YOUR OWN. PLEASE CONTACT YOUR LOCAL LAW ENFORCEMENT AGENCY.” It also recommends taking no individual action with regard to any property listed on the site.
Do not confront the seller. Do not negotiate the price down because of it. Do not take the car to sort it out later. Verify the report is active, then let the police handle it.
A process that actually fits a lot
Putting it together, for someone buying more than one car a month:
- Make the VIN check a step in buying, not a step in buying suspicious cars. The cars that catch you out are the ones that looked fine.
- Use NICB VINCheck where the five searches do the most good, which is on the short list you are seriously bidding on, not on the whole run list.
- Run the dealer-channel history report on everything else. As a licensed dealer you have access to sources your retail buyers do not.
- Check whether your state publishes a public stolen property search, and use it, because it is the only free check drawing on law enforcement records.
- Treat a late-model car with no lienholder in its history as a question, not a bargain.
- Compare the VIN on the car against the VIN on the paperwork, character by character. A clean result against the wrong number is not a result.
- Keep every check you ran, attached to the car you ran it on. A report in your email is not a record you can produce in two years.
That last point is the one people skip, and it is the one that matters when someone asks what you knew and when. Clean title meaning covers the same gap on the title side: the document tells you what to go and check, not what to conclude.
Where this shows up in your margin
Most used-car risks cost you part of a deal. This one can cost you all of it, plus the transport, plus the reconditioning you had already paid for, plus whatever you owe a customer who bought it from you.
Which is why the cost of checking is the wrong thing to optimise. A history report on every car is a line item measured in tens of dollars. One confiscated vehicle is measured in thousands, and it does not show up as a bad month, it shows up as a missing car.
Deelary keeps every cost attached to the individual vehicle, including the checks you ran on it, so the real margin and days held per car stay honest and the paperwork trail stays with the car rather than in your inbox.
Frequently asked questions
How do I check if a car is stolen for free?
Run the VIN through the NICB's free VINCheck, then check whether your state publishes a public stolen property search, as Florida does through the FDLE's public access system. Both are worth doing and neither is conclusive. NICB states that VINCheck does not query law enforcement records, and Florida states its site contains only records that agencies have selected for public access.
Is NICB VINCheck enough on its own?
The NICB says it is not. Its own wording: while VINCheck is a useful free tool, it is not a comprehensive vehicle history report and should not be relied upon when purchasing a vehicle. It cross-references the VIN against participating insurers' theft and salvage records only, and the NICB strongly encourages the public to engage in additional due diligence prior to purchasing a vehicle.
How many free VIN checks can I run in a day?
Five. The NICB states you can conduct a maximum of five searches within a 24-hour period per IP address. The limit is per IP address rather than per person, so a dealership where several people check from the same office network shares one allowance of five.
Can a stolen car have a clean title?
Yes, and that is the point of vehicle cloning. The NICB describes the method: copy a VIN from a legally owned vehicle, create a counterfeit VIN tag and fake ownership documents, steal a matching vehicle, swap the tag, and sell it. The federal guidance notes that in one such operation law enforcement could not detect the vehicles' true stolen status because the paperwork showed legitimate VINs.
What happens if I buy a cloned car?
You lose both the car and the money. The NICB states it plainly: the vehicle is never legally yours, even if you paid for it, and if the VIN switch is discovered the car will be confiscated and returned to its rightful owner or to the insurance company that paid a theft claim on it.
What should I do if a VIN comes back as stolen?
Stop and call the police, not the seller. The NICB says that if a vehicle is reported as stolen but not recovered you should verify with the NICB or your local law enforcement agency that a stolen vehicle report is active before taking any action regarding the vehicle. Florida's FDLE puts it in capitals: do not take action on your own, contact your local law enforcement agency.