How to Geotag Old Photos Without GPS Data
Three honest methods for adding location to old photos with no GPS metadata — manual map tagging, AI visual recognition, and GPS-track matching. Compared side-by-side.
If your photo library contains thousands of pre-2010 camera shots, scanned family albums, or downloaded images that lost their EXIF along the way, you've already noticed: the map view is empty. The photos exist, they have timestamps, but no coordinates. This is solvable in three different ways. The right one depends on how many photos you have and how much of your own time you want to spend.
TL;DR: For a single mystery photo you genuinely don't recognize, try a visual-search tool like FindPicLocation. For a whole library of unknown photos, use a library-aware tool like RetroTagr or digiKam. For trips where you happened to be recording a GPS track on a watch or phone, automatic track-matching is the most accurate option — but it only works if the track exists.
Why old photos lack GPS in the first place
Three common situations leave photos with no location data:
- Pre-2010 digital cameras — most consumer cameras didn't ship with a GPS chip until the early 2010s. A JPEG from a 2007 point-and-shoot has timestamps but no coordinates because the camera literally couldn't record them.
- Scanned film, prints, and slides — when you digitize an album on a flatbed or via a scanning service, the scan inherits the scanner's "now" timestamp and has no GPS at all. The original analog photo never had any metadata to begin with.
- Stripped re-saves — Instagram, WhatsApp, email, and most messaging apps remove EXIF (including GPS) when they re-encode an image. If you saved a photo someone sent you, it likely arrived without coordinates.
The good news: in all three cases, the visual content is intact. That's what makes recognition-based geotagging possible.
Method 1 — Manual map-based tagging
The most established tools let you open a photo, drag a pin on a map, and write the coordinates into the file's EXIF. You name the place; the tool stores it.
- digiKam — free, open-source, cross-platform (Mac, Windows, Linux). The built-in Geolocation Editor uses OpenStreetMap and can geotag photos individually or batch-assign from a GPS track. It's the most powerful free option. Trade-off: it's a full photo manager — you import your library into it before tagging.
- GeoSetter — Windows-only, free, focused purely on metadata editing. Lighter than digiKam if you just want to tag-and-export and don't want a new photo manager.
- Apple Photos (Mac) — Get Info → Add Location. Works for a few photos at a time. Not great for batch tagging hundreds.
- Lightroom Classic — has a Map module that lets you drop pins. Paid subscription, but if you already use Lightroom this is the lowest-friction option.
These tools all do the same job: you supply the location, they write the GPS. Choose based on which platform you're on and whether you want a free option (digiKam, GeoSetter) or a paid one (Lightroom Classic).
Method 2 — AI-based visual recognition
A different category: tools that look at the photo and propose a location based on what's visible. Two players in this space, with different shapes:
- FindPicLocation is built for one-off lookups. You upload a single mystery photo, it runs visual search across multiple models, and returns a likely location. Pricing is credit-based — about $9/month for monthly Pro (~500 credits) and $64/year for annual, with Deep Search using 20 credits per photo. Great when you genuinely don't know where a photo was taken and need a careful guess.
- RetroTagr is built for whole libraries. You import an entire collection of photos that need locations and the AI batch-suggests for all of them. You can accept, edit, or reject each suggestion, and the result writes back to your photo library with standard EXIF GPS. Free tier covers the first 100 photos and 5 AI suggestions; paid tiers add storage and credit packs.
If your problem is this one mystery photo, FindPicLocation is the right tool. If your problem is a thousand photos and I don't want to click a pin on a map a thousand times, RetroTagr is the right shape.
What AI can and can't do, honestly: famous landmarks get street-level accuracy. Distinctive but non-famous places — a specific old town main street, a recognizable mountain ridge — usually land within the right town or region. Indoor shots, generic landscapes, and sky-only photos can't reliably be located better than "country-level" and should be tagged manually using context you remember (trip dates, the album the photo was in, who's in it). RetroTagr flags these low-confidence results so you can skim past them.
Method 3 — GPS track matching
If you happened to be recording a GPS track on a watch (Garmin, Apple Watch, Wahoo) or a phone app (Strava, GPS Logger) on the day a photo was taken, you can match the photo's timestamp against the track and assign the coordinate from where you were standing at that minute. This is the most accurate method by far — it's literal GPS, not inference.
digiKam supports this out of the box (Import GPX → match photos by timestamp). The catch: it only works for trips where you remembered to start the recording.
Which method should you use?
| Situation | Best fit | | -------------------------------------------------- | --------------------------------------- | | One mystery photo, no idea where it is | FindPicLocation (one-off visual search) | | Whole library of pre-2010 camera photos | RetroTagr (AI library tool) | | Scanned family album, decades of photos | RetroTagr (AI library tool) | | Hiking / cycling trip with a recorded GPS track | digiKam + GPX import (track matching) | | Tech-comfortable, want zero cost, willing to learn | digiKam (manual + tracks) | | Already in the Adobe ecosystem | Lightroom Classic Map module | | Windows-only, just want fast manual tagging | GeoSetter |
A common workflow in practice: run the library through RetroTagr's AI first to get the easy wins, then manually tag the low-confidence remainder using GeoSetter or digiKam.
Step-by-step: tagging a library with RetroTagr
This is the AI workflow if you want to follow along.
Find the photos that need locations
In Apple Photos, create a Smart Album with the rule "Photo > Has GPS > is not true" — that's your starting set. In Lightroom, use the Map module's filter "no GPS data". For loose folder structures, ExifTool can list which files have no GPS tags in one command.
Upload to RetroTagr
Drag the folder onto the RetroTagr dashboard, or use Import → From Apple Photos for a direct sync. Uploads run in the background — you can keep using your photo library while RetroTagr processes the batch.
Let the AI propose locations
Each photo is analyzed for visual cues — landmarks, signage, vegetation, terrain, vehicles. A coordinate comes back with a confidence score. Most photos take 3-5 seconds; large batches process in parallel.
Review and accept
For each photo: accept the AI's pin, drag it to refine, or reject and tag manually using the built-in map. The low-confidence results are visually flagged so you can skim past obvious-manual-cases without clicking through every one.
Export back to your library
When you're done with a batch, export. RetroTagr writes standard EXIF GPS (latitude, longitude, optional altitude) into the JPEG/HEIC/RAW. Re-import into Apple Photos / Google Photos / Lightroom — the locations now show up on every map view, every "places" album, every search.
That's it. Five steps; the AI does the bulk of the work, you keep editorial control on every photo before it's written back.