How to Find and Delete Duplicate Files in Google Drive

Native automatic duplicate finder

Not currently provided by Google Drive

Best free starting point

Storage view, ownership filter and name-based review

Most important quota rule

Only files you own count toward personal Google Drive storage

Trash behavior

Drive Trash keeps files recoverable for 30 days but continues to use storage

Automated cloud scan

Conditional option for large libraries after permissions and results are reviewed

Google Photos

A separate library; Drive cleanup does not automatically clean Photos

Evidence checked August 22, 2026. This guide is based on current Google documentation, Google Drive API documentation and the published product, privacy and licensing information for Cloud Duplicate Finder. SoftRadar24 did not perform destructive testing inside private Google accounts, and we do not claim that a third-party scan can identify every kind of duplicate.

This distinction changes the correct tool. Name sorting is useful for obvious copies; exact-content scanning is better for renamed binary files; version history is the right place for revisions; visual review is required for similar images; and Google Photos needs its own workflow.

1. Open Google storage usage details. Identify how much space belongs to Drive, Gmail, Photos, Trash and backups. If Photos or Gmail dominates, a Drive duplicate scan cannot solve most of the shortage.
2. Open the Storage view in Google Drive. Google lists storage-using Drive files from largest to smallest. Review the largest owned videos, archives and disk images before spending time on thousands of small documents.
3. Limit the review to files you own. Use Drive’s advanced search or Owner filter. Files owned by someone else can clutter “Shared with me,” but they do not consume your personal Drive quota.
4. Check Trash, backups and hidden app data. Google specifically identifies these as separate storage sources. Trash still counts; deleting ordinary visible files will not remove a large WhatsApp backup or hidden app data.
5. Record the starting number. Note the Drive storage total before cleanup. Compare it again only after the chosen files are permanently deleted and Google has had time to update the account.

1. Start in the relevant folder or My Drive. Do not begin with “Shared with me” if the goal is to recover personal storage, because non-owned files do not count toward that quota.
2. Switch to list layout and sort by Name. Look for repeated titles and common copy labels such as “(1)” or “Copy of.” Google itself recommends these as manual clues.
3. Narrow the list with search filters. Use file type, owner, location and modified-date filters. Searching the exact suspected title can bring separated copies into one result list.
4. Inspect each suspected pair. Open File information and compare owner, location, size, created time, modified time and activity. Preview or open both files; matching size alone is not proof of matching content.
5. Identify the retained copy. Prefer the file in the intended folder with the required sharing, comments, links and ownership. “Newest” is not automatically better, and “oldest” is not automatically the original that should survive.
6. Move only the confirmed redundant item to Trash. Do not empty Trash in the same session. First verify the retained file from the Drive web interface and any connected device.
7. Review the result after synchronization. Confirm that shared links, collaborators and desktop folders still point to the retained file. Process another small batch only after the first one behaves as expected.
8. Permanently delete when the recovery period is no longer needed. Trash items continue to consume Google storage. After a large permanent deletion, allow up to 48–72 hours for the storage total to update.

Ownership is especially important. If an item does not belong to you, removing it may make the Drive interface look cleaner while reclaiming no personal quota. Conversely, a file you uploaded into another person’s ordinary shared folder can remain yours and continue using your storage.

1. Confirm that the account is suitable. Use your own personal account unless the relevant administrator has approved a managed account. CDF’s FAQ lists Google Workspace support, while its terms also contain restrictions on commercial use and scanning another individual’s content; organizations should obtain clarification before proceeding.
2. Read the privacy policy and current terms. The vendor says it does not store Google passwords or files, but it does store account identity information and metadata for duplicate Google Drive files. Decide whether that data handling is acceptable before connecting Drive.
3. Create a CDF account and connect Google Drive through Google’s authorization screen. Confirm the selected Google account and read every requested permission. OAuth avoids sharing the Google password, but the granted scope can still authorize file operations.
4. Select one noncritical test folder. Do not begin with the entire account, a shared team folder or irreplaceable family archives. A focused folder makes false assumptions easier to detect.
5. Choose the correct scan type. Use All Files or a suitable Custom scan for exact copies. Activate Similar Photo Finder only when visually related images are intentionally being reviewed; it is a different problem with a higher judgment requirement.
6. Run the free scan and inspect the groups before buying. The vendor says scanning and viewing detected duplicates are free, while moving or deleting them is a paid action. Verify that the results address the real problem before paying.
7. Review the proposed retained file in every group. Automatic rules such as largest, smallest, newest and oldest are selection shortcuts, not evidence of importance. Check ownership, location, sharing, quality and actual content.
8. Use Move to Folder for uncertain groups. A dedicated review folder preserves active access and provides a second inspection stage. Use Move to Trash only for files already confirmed as redundant.
9. Verify Drive and connected computers. Open the retained files, test important shared links where possible and let Drive for desktop finish synchronizing before processing another batch.
10. Revoke access when the job is complete. Remove the linked application from Google Account connections if ongoing access is unnecessary. Revocation stops future API access; it does not automatically erase data already held by the vendor, so use its account or support deletion process separately if required.

If two exact copies are equally valid, choose one authoritative location and document it before removal. Reducing duplication without establishing a clear source of truth often allows the same copies to return during the next upload, restore or synchronization cycle.

• Keep intentional edits. A cropped, color-corrected or annotated image may serve a different purpose from the original.
• Compare dimensions and file format. A small preview and a full-resolution source can look the same on screen while having very different archival value.
• Check people and moments carefully. Burst photos can differ by expression, focus or motion even when the scene is almost identical.
• Do not generalize the photo feature to videos or documents. The vendor publishes a similar-photo mode; that is not evidence of perceptual matching for every file category.

Permission and Account Checklist

• Confirm the application name and selected Google account on the consent screen.
• Read the requested permissions rather than relying on a generic “secure” label.
• Do not connect a work or school account without the required administrator approval.
• Use a unique password and multi-factor authentication for the separate CDF account where available.
• Begin with one noncritical folder and review the free results.
• Revoke the Google connection after cleanup when continuing access is unnecessary.
• Contact the vendor separately if you also want retained account information and duplicate-file metadata deleted.

A Safer Deletion Sequence

1. Move uncertain candidates to a review folder rather than Trash.
2. Open the retained copy and verify its owner, folder and sharing.
3. Move confirmed redundant files to Trash in a small batch.
4. Let Drive for desktop synchronize and inspect one connected device.
5. Keep the Trash recovery window for as long as the storage situation allows.
6. Permanently delete only after verification, then wait for Google’s storage accounting to update.

• Deleting by filename alone. Drive filenames are not required to be unique, and a renamed exact copy can escape a name-based search.
• Deleting by file size alone. Equal size does not prove equal content.
• Assuming every visible file uses your quota. Only owned files count toward personal Drive storage; removing non-owned shared items may reclaim nothing.
• Confusing shortcuts with copies. A shortcut is a reference to a target, not an independent stored duplicate.
• Confusing revisions with separate files. Use Manage versions or Docs Editors version history for revision problems.
• Automatically keeping the newest file. A recent item can be a compressed export, restored copy or upload to the wrong folder.
• Bulk-deleting similar photos. Similarity thresholds create candidates, not a guarantee that one image has no distinct value.
• Emptying Trash immediately. This removes the normal recovery path before synchronization and sharing have been checked.
• Assuming Drive and Google Photos are one library. The services share account storage but their file changes and deletions are separate.
• Granting third-party access without reading the consent screen. OAuth protects the password exchange; it does not make file-management permissions harmless.
• Scanning a managed account without approval. Workspace policy, shared-drive ownership, records requirements and vendor licensing can all apply.
• Buying before reviewing the free results. CDF says scanning is free; confirm that its groups match the actual problem before paying for move or delete actions.

Whichever method you choose, separate exact copies from similar or revised content, keep the file with the correct ownership and collaboration context, test one narrow folder, and delay permanent deletion. The objective is not to produce the largest “space recovered” number—it is to recover useful storage without removing the one copy that people, links, synchronized devices or future recovery still depend on.