How accurate is an AirTag's location?
An AirTag location is not a measurement of the tag. It is the reported position of whichever passing iPhone heard it, so the accuracy you get is that phone's accuracy. Outdoors in a city that usually lands on the right building; indoors or between tall buildings it can be looser. Precision Finding is a separate, in-the-room feature.
Whose location is an AirTag report, actually?
Not the tag's. An AirTag has no GPS receiver and no way to determine its own position. When a stranger's iPhone hears the tag's Bluetooth advertisement, that phone works out where it is, encrypts its own coordinates, and uploads them. What you eventually decrypt is a statement about the phone.
The implied claim is a modest one: this tag was within Bluetooth range of a device that believed it was here, at this time. Bluetooth range is short, so that is a genuinely useful claim, but it is a proximity statement, not a fix on the tag.
Everything about AirTag accuracy follows from this. The point inherits the finder phone's positioning quality, plus however far apart the two of them were when the phone listened.
What decides how good the relaying phone's position is?
The same things that decide how well the blue dot works on your own phone. Outdoors with a clear view of the sky, a modern phone knows where it is well. Indoors it falls back on nearby Wi-Fi and cellular signals, which are coarser.
- Sky view: an open street or a car park gives the phone satellites to work with; a warehouse interior does not.
- Tall buildings: signals bounce off glass and stone, which is why downtown points sometimes land across the road.
- Indoors: position comes largely from Wi-Fi and cell towers, so it tends toward the building rather than a spot inside it.
- Movement: a phone in a moving car reports where it was at that instant, which may be past the tag by the time the report lands.
- Distance: the phone could be at the far end of Bluetooth range from your tag, and nothing in the report records that offset.
None of this is adjustable and none of it is visible in the report. You get a latitude, a longitude, and a timestamp, with no confidence figure attached.
Isn't Precision Finding accurate to the centimetre?
Precision Finding is a different mechanism that answers a different question. It uses ultra-wideband radio between a supported iPhone and the tag itself to give you a direction and a distance, walking you in the last stretch: under a sofa, behind a shelf, in the wrong suitcase.
It only works when you are already in the room. Ultra-wideband is short-range, needs your own phone, and produces nothing at all otherwise. It never contributes to the location reports that arrive from the network, so it does not sharpen a single point in your history.
Two different tools, then: the network tells you which building to drive to, and Precision Finding takes over once you are standing in it. Confusing them is why people expect metre-level points from a shipment on the far side of a country.
How accurate should you expect a point to be?
In ranges rather than figures, because nothing in a report tells you how good that particular report was. What follows is what people typically see, and it is worth reading as a shape rather than a specification.
| Where the tag is | What the point tends to be worth | What it lets you conclude |
|---|---|---|
| Open street in a town or city | Usually the right building, or the right side of the road | The address or block, enough to name a place |
| Dense downtown, tall buildings | Can land a building or two off, sometimes across the street | The block, not the doorway |
| Inside a warehouse or terminal | Often the building envelope rather than a spot within it | Which facility, not which bay |
| Rural road or open country | Coarse and infrequent, whenever a vehicle passes | Roughly which stretch of road, and when |
| Moving vehicle | Wherever the relaying phone was at that instant | Direction and progress, not a precise position |
| Underground or sealed in metal | Few points, or none at all | Nothing, until it surfaces |
Why is a point that is a building off still worth having?
Because of the size of the question you are usually asking. Nobody arguing with a moving company needs to know which shelf a crate is on. They need to know it was in Rotterdam on the 14th when the paperwork says it was still in Bristol.
At that scale, a point that is a hundred paces out is exactly as convincing as one that is perfect. The error that matters is the kind that would change the answer: a different city, a different warehouse, a different day. Points are rarely wrong like that.
- Was the shipment at the depot, or had it already left?
- Did the bag make the transfer, or stay behind at the first airport?
- How long did the container sit in the yard before it moved?
- Did the item ever reach the address the courier claims it reached?
Every one of those survives a loose point. None of them survives having no record at all, which is what Find My leaves you with the moment the next report overwrites the last.
What should you do when a point looks obviously wrong?
Read it as the phone's position and most oddities explain themselves. A point in the middle of a motorway usually means a passing driver's phone relayed the sighting. A point in the building next door means the finder was in the building next door.
- Trust the cluster, not the outlier. Many reports from roughly the same place over hours is a far stronger claim than any single point.
- Check the timestamp before the coordinates. An old report shown as current is the more common misreading.
- Treat a lone point far from the rest as one phone with a bad position, until a second point agrees with it.
- Remember the tag can be anywhere within Bluetooth range of the reported spot, in any direction.
This is also why a history is more defensible than a pin. One point is an assertion; a run of them across a day is a pattern, and a pattern survives the one report that landed across the street.
Frequently asked questions
How accurate is an AirTag's location?
As accurate as the phone that reported it. An AirTag cannot measure its own position; each point is the location of a passing Apple device that heard the tag over Bluetooth. Outdoors that usually lands on the right building or block, while indoors it tends to resolve to the building rather than a spot inside it.
Why is my AirTag showing the wrong location?
Most often it is not wrong, it is old or offset. Find My shows the most recent report whenever it arrived, so a stale point looks like a claim about now. An offset point usually means the phone that heard your tag was across the street or had a poor position of its own indoors.
Is Precision Finding more accurate than the Find My network?
Yes, but only in the last few paces. Precision Finding uses ultra-wideband between your own supported iPhone and the tag to give direction and distance once you are close. It is short-range, needs you to be there, and contributes nothing to the location reports stored in a history.
Can an AirTag tell me which room something is in?
Not from the network. Reports typically resolve to a building rather than a room, because the relaying phone is usually indoors and working from Wi-Fi and cell signals. Precision Finding can walk you to the object once you are inside the building, but it needs you physically present.
Does an AirTag have GPS?
No. There is no GPS receiver, no SIM and no cellular data in an AirTag. Every position you see came from another Apple device's own positioning. That is why accuracy varies with where the tag is rather than with anything about the tag, and why there are no coverage charges anywhere in the world.
Is AirTag location accurate enough for an insurance claim?
Usually, because claims turn on which place and which day rather than which metre. A timestamped run of points showing an item at a depot for a week answers the question a claims desk is asking. Being a building off does not weaken that; having no record at all does.
Why do two AirTags in the same bag show slightly different locations?
Because they were heard by different phones, or by the same phone at different moments. Each report carries whatever position the finder device had at that instant, so two tags travelling together will often land a short distance apart. The spread between them is a rough picture of the uncertainty.
Start keeping the route
AirTag History polls your AirTags every 5 minutes and keeps the timeline Find My overwrites. Free for 7 days, and your history starts the moment you connect, not before, so connect ahead of the move.