Your Dog's Microchip Works. The Phone Number Might Not.

A microchip is not a tracker, and it is not a promise. It is one number under the skin that only pays off if a stranger can follow it back to your phone. Researchers have measured every step of that handoff, and the results are lopsided: the implant almost never fails. The paperwork does.

Where the chain actually breaks

The study everyone still quotes is Lord and colleagues in the Journal of the American Veterinary Medical Association, 2009, summarised by Ohio State's College of Veterinary Medicine: 53 US shelters logged every animal arriving with a chip. Median return-to-owner rates ran 2.4 times higher for microchipped dogs than for strays generally — 52.2% against 21.9%.

Then they logged why the rest failed, and the pattern is the point. The single largest cause was a wrong or disconnected phone number. Next came owners who never returned a call or a letter, then chips registered in a database other than the manufacturer's, then chips never registered at all. Those four causes account for 86.7% of the listed failures — and every one of them sits downstream of a scan that already worked.

The scanner is rarely the problem. In the recorded failures the chip was found and read. What broke afterwards was a phone number, a registry choice, or a form nobody finished.

"Get your dog chipped" is the easy half, and it is the half the whole internet repeats. The half that decides whether he comes home is administrative, free, and almost nobody does it twice.

Read the number like a serial number

Your dog's chip number is not arbitrary. WSAVA's microchip identification guidelines set out the ISO standard: a compliant chip carries a 15-digit number whose first three digits are a manufacturer or country code, transmitting at 134.2 kHz. Older US chips are different animals — a 10-character alphanumeric code at 125 kHz, or a 128 kHz variant. AAHA's lookup tool accepts 9, 10 or 15 digits, which tells you how many formats are still circulating. It also flags one specific ambiguity: numbers beginning 900 often cannot be traced to a single company, so the tool returns several.

The registry is a market, not a database

There is no national registry. AAHA's lookup lists roughly three dozen separate companies holding chip-to-owner records, and nothing obliges the chip inside your dog to be registered with the company that made it. That is the 17.2% bucket, in one sentence.

Worse, registries are businesses, and businesses close. AAHA's tool currently carries a notice that Save This Life ceased operations and was disconnected from the lookup, advising owners to get their number from their vet and re-register elsewhere. Nobody emails you when that happens.

A registration is an account at a company, not a permanent filing. The number in your dog lasts his whole life. The record pointing it back at you does not.

What the lookup will and will not do

The free lookup is the right first step, and it is narrower than most people assume.

• It will name the registries associated with your chip number, most recent first.
• It will not show owner contact details, or let you edit them.
• It will list companies that may have sold an unregistered chip, so a shelter can request trackback information on whoever bought it.
• It will not stand in for phoning the registry — AAHA states plainly that it does not manage registries and cannot resolve chip problems itself.

The tool tells you where to call. You still have to call.

The scan is not guaranteed either

The US never adopted a national standard, so three frequencies circulate at once — 125 kHz still predominant, 128 kHz, and the ISO 134.2 kHz — plus encrypted and unencrypted versions of the 125 kHz protocol. Ohio State's scanner trials concluded that no single scanner reads every chip with 100% sensitivity, and named the conditions that degrade a read:

• A metal collar left on during the scan
• Scanning beside computers, metal tables or fluorescent lighting
• A tired scanner battery
• One pass instead of several

If your dog is ever handed in somewhere, "we scanned him and there was nothing" is not the same statement as "he has no chip." Ask whether the scanner was a universal one, and whether he was scanned more than once.

The evidence limit worth naming

Those percentages are the best evidence available, and they are also the only evidence available. The AVMA was still citing that same 2009 study in 2024, because no comparable US replication has been published since. The data behind it was gathered when fewer companies held these records than hold them now. Read the figures as a map of where failures cluster, not as a current forecast — and note that every change since points the same way: more registries, more fragments of the same record.

The ten-minute check

1. Find the number. Vet record, adoption paperwork, or ask for a scan at the next appointment.
2. Run it through the lookup and write down every registry it returns, not only the first.
3. Contact each one. Confirm the number on file is the phone in your pocket today.
4. Add a second contact who is not you — a partner, a neighbour, anyone who answers.
5. Hang an engraved tag on the collar. Visual ID needs no scanner, no registry and no phone call.
6. Answer unknown numbers while a dog is missing. A quarter of the failed reunions were owners who never picked up.

None of that costs anything, and none of it is what the chip was sold to you as. A chip is recovery infrastructure, not live tracking — it reports your dog's location exactly once, after somebody else already has him. If you want to know where he is right now, that is a different device entirely.