What the GPS data in your photos reveals (and how precise it really is)

When a phone with location services enabled takes a picture, it writes a small block of GPS tags into the file next to the image. That block is easy to overlook because no viewer shows it by default, but it is also the most consequential piece of metadata a photo can carry. This guide walks through what the tags actually contain, how precise the numbers are in practice, and what a handful of geotagged photos can reveal when someone looks at them together.
Where the coordinates live
EXIF stores location in its own directory, the GPS IFD, which the main image directory points to through tag 0x8825. Inside it, latitude and longitude are written as three rational numbers each — degrees, minutes and seconds — with separate reference tags for N/S and E/W. A typical phone file also contains GPSAltitude with a flag for above or below sea level, GPSTimeStamp and GPSDateStamp taken from the satellite clock, and often GPSImgDirection, the compass heading the camera was pointing at.
Newer devices add GPSSpeed with its unit, GPSHPositioningError (the receiver's own estimate of its horizontal error in metres) and GPSProcessingMethod, which can say whether the fix came from satellites, Wi-Fi or cell towers. None of this is hidden in any technical sense: it is plain, documented structure that any EXIF reader decodes in milliseconds.
How precise the numbers are
One second of latitude is about 31 metres on the ground, and phones write the seconds value with two or more decimal places, so the format itself resolves to well under a metre. The real limit is the receiver. Outdoors with a clear sky, a modern phone combining GPS, Galileo and GLONASS with Wi-Fi and cell corrections lands within roughly 3 to 10 metres of the true position. Between tall buildings or indoors the error grows to tens of metres, and the positioning-error tag, when present, tells you what the device itself believed.
Ten metres is enough to identify a building, and often the entrance. Altitude is less reliable — barometer-assisted phones are usually within a few metres, older ones can be off by fifty — but it still separates a ground-floor flat from a tenth-floor one. Heading from the magnetometer is typically accurate to 10–20 degrees, which is enough to tell which window a picture was taken from.
More than one point: time and direction
The GPS block carries its own timestamp in UTC, while DateTimeOriginal in the Exif directory is local time. Comparing the two gives the time zone the photo was taken in even when the OffsetTime tags are missing. Speed tells whether you were walking, cycling or in a car. Direction, combined with the coordinates, points at the subject rather than the photographer: the door you faced, the neighbour's house, the school gate.
A single photo is a point. A folder is a pattern. Evening and weekend photos cluster at home; weekday-morning ones at a school or a workplace; the same café on Saturdays; the same park loop after dinner. Anyone who receives a batch of unstripped pictures — a buyer on a marketplace, a forum, a former partner — can reconstruct that routine without any special skill, just by sorting by time and dropping the points on a map.
Who actually reads it
Reading GPS tags needs nothing exotic. Windows shows latitude and longitude in the Details tab of a file's properties; macOS Preview draws a small map in its inspector; Android and iOS gallery apps show a location card under the picture. Command-line tools such as ExifTool dump everything, and browser extensions do the same for images on web pages. If a file with intact metadata reaches someone, assume they can see where it was taken.
Large platforms usually strip the block from the copy they publish, but they read it first, and a surprising number of smaller sites, forums, email clients and file-sharing links do not touch it at all. Our guide on which platforms strip EXIF goes through the details; the short version is that you cannot rely on the recipient's side.
What to do about it
The cleanest fix is to stop writing the tags in the first place. On iPhone, Settings → Privacy & Security → Location Services → Camera → Never; on most Android phones the camera app has a “Save location” or “Location tags” switch. Both systems also offer a per-share option: the iOS share sheet has Options → Location, and Google Photos can hide location data in links. Those switches only affect new photos and new shares, though; the pictures already on your phone keep their tags.
For everything that already exists, strip the block before sharing. Drop the folder into ExifMop, look for the red pins, and either remove only the GPS group or wipe the metadata entirely. The removal is lossless, so there is no reason to keep an untouched copy for quality's sake — keep it only if you want the location for your own archive.
The takeaway
A photo's GPS block resolves to within a few metres, carries time and heading as well, and a series of photos maps your routine. Turn off location for the camera, and strip the tags from anything already taken before it leaves your device.
More guides

Which apps and platforms strip EXIF — and which don't
Social networks mostly strip metadata from what they publish, messengers depend on how you send, and email and cloud drives keep everything. What is known as of 2026, and how to check for yourself.

Lossless EXIF removal vs re-encoding: why the mop matters
Metadata sits in its own segments and chunks, so it can be cut out without touching a pixel. Re-encoding does the opposite: it recompresses the picture, can shift colours and drops the orientation you relied on.