How to Find Duplicate Photos on Windows

Most duplicate photos are not identical files. That is why sorting by file size does not find them, and why the folder you were sure you had cleaned up still has the same holiday in it four times.

The short answer

1. Open Photo Organizer in PhotoTray.
2. Point it at the folders your photos actually live in.
3. Run a duplicate scan.
4. Review each set side by side and choose what to keep.

Nothing is deleted automatically, and nothing is uploaded.

Why your duplicates are invisible to most tools

The naive approach compares file sizes, or hashes each file and looks for matches. That finds exact copies — the result of a straight file copy — and misses almost everything else.

Real photo libraries accumulate duplicates that differ at the byte level while being obviously the same picture:

• The same photo imported at full size once and as a resized copy another time.
• A JPG that has been re-saved by a different app, changing the compression.
• The HEIC original and the JPG you converted from it.
• A photo that got rotated or had its metadata edited.
• The same image from a phone backup and from a cloud download, saved with different filenames and different EXIF.

To a checksum, every one of those pairs is two unrelated files.

What perceptual comparison does instead

The approach that works is to compare what the image looks like.

Each photo is reduced to a compact fingerprint derived from its visual structure — the broad pattern of light and dark, largely independent of resolution, file format and compression. Two files that look the same produce near-identical fingerprints even if their bytes have nothing in common.

That catches the resized copy, the re-saved JPG, the converted HEIC and the rotated version, which is most of what is actually cluttering the drive.

Why nothing is deleted for you

This is the part where photo tools cause damage, so it is worth being explicit.

A duplicate set is frequently not a set of unwanted copies. It is often:

• The original and the one you edited, which looks nearly the same and is the one you want.
• A full-resolution file and a small copy that is already embedded in a document or a shared album.
• Two frames from a burst that are genuinely different photographs of the same moment.

An automatic "keep the largest" rule gets all three of those wrong. So Photo Organizer shows each set side by side, with sizes, dates and folders, and waits.

This is also why the tool is deliberately excluded from PhotoTray's folder rules. Every other utility in the app can be scheduled to run unattended; this one cannot, because deciding which copy of a photo to keep needs a person.

Where to point it

Duplicates cluster in predictable places:

• Pictures, and every "New folder" inside it.
• Downloads, which quietly accumulates re-downloaded images.
• Old phone backup folders from devices you no longer own.
• Camera import folders, especially if you have ever imported the same card twice.
• Anywhere you copied photos "just to be safe" before a reinstall.

Scan them together rather than one at a time. The interesting duplicates are usually the ones spread across folders, not within one.

A word about cloud folders

If your photos live in OneDrive, iCloud Photos, Dropbox or Google Drive, be careful — and this applies to any tool, not just this one.

Scanning is safe: it only reads. But deleting or moving a file inside a synced folder propagates that change to every device on the account, immediately, and sometimes irreversibly. Cleaning up duplicates on your PC can quietly empty them off your phone too.

PhotoTray detects those folders and warns you explicitly before it changes anything inside one. That warning is worth reading rather than clicking through.

Rescans are cheap

The first pass over a large library takes a while, because every photo has to be read and fingerprinted. Results are kept locally afterwards, so a second scan only looks at what has changed.

That makes this something you can run every few months rather than a one-off afternoon project — which is the difference between a library that stays tidy and one that gets cleaned once and immediately starts drifting again.