calibre Duplicate Books Not Detected When Adding: How to Fix It

When calibre adds a second record for a book already in your library, the program may not be malfunctioning. Duplicate handling during import depends heavily on the title and author metadata calibre reads, the Add books action you use, and your selected automerge behavior. A changed subtitle, reversed author name, missing author, filename-based metadata, or different edition can make two related files appear to be separate books. Use the controlled tests below to identify the cause before reinstalling calibre, deleting records, or modifying your main library.

1. Confirm the Symptom With a Small Safe Test

Start by confirming exactly what calibre is doing. A duplicate can mean either two records for the same logical book or two legitimate editions that happen to share a title. Those situations require different fixes.

1.1 Compare the existing and incoming metadata

Find the existing book in calibre and open Edit metadata. Write down or copy its Title, Author(s), and identifiers. Then inspect the incoming file by adding it to a temporary library, as described later, or by using calibre's metadata tools.

Compare the values character by character. Common differences include:

• A subtitle appearing in one title but not the other
• An edition label such as Revised Edition or Second Edition
• Initials in one author name and full names in the other
• Author names in different orders
• An editor, translator, or narrator listed as the author
• Unknown or blank author metadata
• Punctuation, bracketed series information, or inconsistent spacing
• Different spellings or accented characters

Do not focus only on the filename. Depending on your Adding books preferences and the file type, calibre may read metadata embedded inside the book instead. A perfectly named file can therefore arrive with an unexpected title or author.

Success looks like this: you identify a specific metadata difference that explains why calibre treated the file as a separate book. Once the difference is known, stop changing unrelated settings and correct the metadata or use the appropriate add-to-record command.

1.2 Run a two-file test

Choose one existing book and make a backup copy of an additional format or test file you are legally permitted to use. Avoid testing with hundreds of books. Add only that file and observe whether calibre displays a duplicate warning, merges it, ignores it, replaces a format, or creates another record.

Repeat the test after making the incoming title and author match the existing record exactly. If matching metadata changes the result, the issue is metadata recognition rather than a damaged installation.

If the test file is simply another format of a known book, select the existing record and use the menu available by right-clicking Add books to add files to selected book records. You can also drag the file onto the Book details panel for the selected record.

Success looks like this: the existing record gains another format, such as EPUB or PDF, without creating another row. Stop troubleshooting once that happens consistently.

2. Check Add Books Settings and Import Behavior

The most important settings for this symptom are under Preferences, Import/export, Adding books. Names and layout can vary slightly between calibre releases, but the relevant areas concern metadata reading, adding actions, automatic adding, and automatic merging.

2.1 Verify where calibre reads metadata

calibre can read metadata from inside supported e-book files or infer it from filenames. If the embedded metadata is poor, reading the file contents may produce a generic title, an incorrect author, or edition text that does not match your existing record. If filename parsing is enabled or used as a fallback, an unsuitable filename pattern can cause similar problems.

In Adding books preferences, review the option concerning reading metadata from file contents. Then examine the filename pattern used to guess the title and author. A common pattern expects a structure similar to Title - Author. Files named in a different order may be interpreted incorrectly.

Test both approaches with one copied file:

1. Import it into a temporary library using embedded metadata.
2. Record the resulting title and author.
3. Remove the test record from the temporary library.
4. Change the metadata-reading option and import the copy again.
5. Keep the method that produces the cleanest, most consistent fields.

Success looks like this: a test import produces the same title and author convention used in your main library. Stop changing the filename expression once several representative files import correctly.

2.2 Review automatic merge behavior

Look for the Adding actions option that controls what happens when books with similar titles and authors are found. Depending on your choice, calibre can merge incoming formats into an existing record, ignore a duplicate format, overwrite an existing format, or place the incoming file in a new record.

Choose conservatively:

• Ignore: useful when the existing copy must be preserved and duplicate formats should be skipped.
• Overwrite: appropriate only when you intentionally want the incoming file to replace the same format already stored.
• New record: preserves both files as separate records, which can look like failed duplicate detection.

If your goal is to combine an EPUB and PDF belonging to the same title, automerge can help when their title and author metadata are sufficiently similar. It is not a substitute for reviewing different editions.

Success looks like this: adding a different format with matching metadata updates the intended record according to the chosen rule. Once the result matches your desired policy, do not continue switching between merge modes.

2.3 Check automatic adding separately

If you use an automatic import folder, inspect the Automatic adding tab or section. Automatic adding has its own duplicate-checking behavior. A duplicate-related checkbox that is disabled can allow files from the watched folder to become new records even though a manual Add books test behaves differently.

Test by placing one copied file in the watched folder. Keep the original outside that folder and wait until calibre finishes processing the file. Compare the result with a manual import of the same file into a temporary library.

Success looks like this: manual and automatic imports follow the duplicate policy you selected. If only automatic adding fails, concentrate on its settings rather than reinstalling calibre.

2.4 Understand format-only duplicates

A second file is not automatically a second book. An EPUB, PDF, MOBI, or AZW3 file can be another format of the same logical title. For a small number of files, the safest method is to select the existing record and choose Add files to selected book records.

If the selected record already contains that format, calibre may replace it, reject it, or create a separate record depending on the action and preference used. Back up any format you need to preserve before allowing replacement.

Conversion does not require creating a second book record. Converting an EPUB to another format normally attaches the output to the same record. If you exported the converted file and later imported it as a new book, calibre must evaluate it as a fresh import and may create another record.

3. Fix Metadata, Identifiers, and Edition Differences

Duplicate detection during Add books should not be treated as an ISBN-only catalog matching system. Title and author metadata are especially important to import and automerge behavior. Identifiers remain valuable for confirming identity, searching the library, and distinguishing editions, but a shared or missing ISBN does not guarantee the desired merge result.

3.1 Normalize titles and authors before bulk import

For a large batch, do not import directly into the main library and hope that every record will match. Use a temporary intake library and clean the following fields first:

• Remove unwanted filename fragments from titles
• Apply a consistent policy for subtitles
• Standardize author names
• Correct records whose author is Unknown
• Move edition information into an appropriate, consistent field or title convention
• Review series names and numbers without treating them as part of the title unnecessarily
• Add reliable identifiers when available

You can edit individual records or use bulk metadata editing for genuinely shared changes. Do not apply one author or title correction to an entire mixed selection.

Success looks like this: records in the intake library use the same naming conventions as the main library and can be compared without interpreting noisy filenames.

3.2 Treat ISBNs and other identifiers as evidence

A missing ISBN is common in personal documents, older files, independently published books, and files with incomplete metadata. Different formats of the same edition may also carry inconsistent identifiers. Conversely, hardcover, paperback, revised, and regional editions may legitimately have different ISBNs.

Before merging, ask:

• Do the title and primary author refer to the same work?
• Is the incoming file the same edition or a meaningfully different edition?
• Does one copy contain revisions, annotations, illustrations, or a different translation?
• Would keeping both formats under one record hide an edition distinction you care about?

If two files are the same work and edition, adding them as formats under one record is usually cleanest. If they represent distinct editions, retaining separate records and labeling each clearly is often safer.

3.3 Merge records created during import

If calibre has already created duplicate rows, select the records that truly represent the same book and use Edit metadata, Merge book records. Review which record should supply the preferred metadata and confirm that the formats you want to keep are present.

Before merging, check for two copies of the same format. A merge may require choosing which EPUB, PDF, or other file should survive. Compare file size, modification date, quality, table of contents, and edition details rather than assuming the larger or newer file is better.

Success looks like this: one record remains, it contains the desired formats, and its title, author, cover, identifiers, and comments are correct. Stop after verifying the merged record opens successfully.

4. Rule Out Library, Plugin, and Operating System Interference

Permissions, security software, and storage problems are less common causes of this exact symptom than metadata or Add books settings. They become relevant when imports behave inconsistently, jobs fail, metadata changes do not persist, or the library database cannot be updated reliably.

4.1 Confirm that you are using the intended library

calibre can manage multiple libraries. Check the active library name and path before testing. A book may appear to be undetected simply because the matching record is in another library or virtual library filters are hiding it.

Clear the search box and exit any virtual library view. Search for the exact title, then search by author and identifier. If you use an intake library, confirm whether you are copying or moving records to the main library rather than importing the source files again.

4.2 Keep the active library on suitable local storage

Do not troubleshoot duplicate handling while the active calibre library is being modified by cloud synchronization or opened from an unreliable network filesystem. File locking and synchronization conflicts can interfere with the database and stored book folders.

Pause synchronization while calibre is open. For diagnostic purposes, create a small library on a normal local folder to which your user account has full read and write access. Avoid placing the test library inside a protected operating system directory.

On Windows, also check whether ransomware protection or antivirus software is blocking calibre from modifying the library folder. On macOS, review privacy permissions if the library is in a protected location. On Linux, verify folder ownership and write permissions.

Success looks like this: imports and metadata changes persist in a local test library without errors. If they do, correct the main library's storage or synchronization arrangement instead of changing duplicate rules.

4.3 Disable import-related plugins for one test

File type and metadata plugins can influence how incoming files are processed. Temporarily disable only plugins that affect import or metadata reading, restart calibre, and repeat the one-book test in a temporary library.

Do not remove every plugin at once. Disable one relevant plugin or a small related group, test, and restore it if behavior does not change. If a plugin is responsible, check its configuration and compatibility before deciding whether to keep it disabled.

The optional Find Duplicates plugin can help locate duplicates after import, especially when you need broader matching choices. It does not replace careful edition review, and it should not be used as permission to merge every similar title automatically.

4.4 Separate device and server behavior from library imports

Sending a book to a connected reader is not the same operation as adding it to the calibre library. Device drivers and reader databases have their own filename, metadata, and replacement limitations. Likewise, downloading through the Content server does not determine how desktop Add books matches incoming records.

Reproduce the problem with the device disconnected and the Content server stopped. If a normal local Add books test works, the core library import is functioning. Investigate the device transfer or remote workflow separately.

5. Use Logs and Debug Output When the Result Is Inconsistent

Logs are useful when the same file produces different results, a job fails, or a plugin raises an error. They are less useful when the cause is an obvious title or author mismatch.

5.1 Review job details first

After an import or metadata operation, open the Jobs area and inspect details for failures or warnings. Save the text before closing calibre if you may need to compare tests.

Look for references to the source file, metadata-reading plugin, database, permissions, or inaccessible paths. A clean job combined with a new record usually points back to matching expectations rather than a technical failure.

5.2 Start the graphical interface in debug mode

Advanced users can close calibre and launch its graphical interface from a terminal with calibre-debug -g. On macOS, command-line tools are stored inside the calibre application bundle unless they have been made available another way.

Reproduce the issue with one non-sensitive test file, then close calibre and review the terminal output. Do not publish logs without checking them for usernames, folder paths, email addresses, server addresses, or book information.

Success looks like this: the debug run identifies an import plugin, unreadable file, permission error, or other repeatable cause. If no error appears and metadata differs, return to metadata normalization rather than pursuing deeper debugging.

6. Run a Clean Temporary Test Before Reinstalling

Reinstalling calibre rarely changes metadata embedded in an e-book, the import preferences stored in your configuration, or the structure of an existing library. A temporary library is faster and safer.

1. Use Switch/create library to create an empty test library in a local folder.
2. Copy, rather than move, two or three representative source files into a separate test folder.
3. Add the first file and correct its title and author.
4. Add another format or copy representing the same book.
5. Observe the duplicate prompt or automerge result.
6. Repeat after changing only one Adding books setting.
7. Record which combination produces the intended behavior.

If the temporary library works, your calibre installation is probably functioning. Compare the main library's active preferences, plugins, metadata, and storage location. If the temporary library fails in exactly the same way, confirm whether the result is expected from the incoming metadata before considering a clean configuration test.

Do not delete the main library, restore its database, or run broad maintenance solely because duplicate records were created. Database restoration is intended for actual database damage and can discard some library-specific information. Back up the library before any repair operation.

7. Quick Fix Checklist

• Confirm that both files represent the same work and edition.
• Compare the exact title and author metadata, not just filenames.
• Check whether calibre reads embedded metadata or guesses from filenames.
• Review the automerge choice under Adding books preferences.
• Check duplicate handling separately if you use automatic adding.
• Add an extra format directly to the selected existing record.
• Use a temporary intake library for large, messy batches.
• Normalize titles, authors, subtitles, and edition labels before copying records.
• Use identifiers as supporting evidence rather than the only matching rule.
• Merge confirmed duplicate records after reviewing overlapping formats.
• Pause cloud synchronization while calibre uses the library.
• Test on local writable storage with import-related plugins disabled.
• Use job details or calibre-debug -g only if behavior remains inconsistent.
• Stop changing settings as soon as a controlled test produces the desired result.

8. Frequently Asked Questions

8.1 Why does calibre create another record for the same book?

The incoming file may have a different title or author, your automerge policy may permit new records, or you may be using the general Add books command instead of adding a format to the selected record. Different editions may also deserve separate records.

8.2 Will adding an ISBN make calibre detect every duplicate?

No. ISBNs help establish identity and distinguish editions, but Add books duplicate handling should not be considered an ISBN-only matching system. Make the title and author consistent, then use identifiers to confirm whether editions are equivalent.

8.3 How do I add a PDF to an existing EPUB record?

Select the existing book, right-click Add books, and choose the option to add files to selected book records. You can also drag the PDF onto the selected book's Book details panel. Verify that both formats appear in the same record afterward.

8.4 Should different editions be merged?

Not automatically. Keep separate records when editions have different translations, revisions, illustrations, pagination, annotations, or identifiers that matter to you. Add an edition label so the records can be distinguished easily.

8.5 Can I find duplicates after importing a large batch?

Yes. Search by title, author, and identifiers, or use a reputable duplicate-finding plugin with carefully selected matching rules. Review every proposed merge when different editions or common titles may be present.

8.6 When should I reinstall calibre?

Reinstall only after a controlled temporary-library test shows a repeatable program failure that is not explained by metadata, settings, plugins, permissions, or storage. If the import completes normally but produces a new record, reinstalling is unlikely to change the matching result.