Fix: Notion CSV Import Creates Duplicate Rows Instead of Updating

Why this happens

Notion's standard Import action (Settings & members → Import, or the "..." menu on a database) is additive by design, it appends every row from the CSV as new pages in the database. It has no concept of "this row already exists, update it", so importing the same file (or an updated version of it) a second time always creates a full second set of rows rather than reconciling with what's already there.

The fix: use "Merge with CSV" instead of "Import"

1. Open the database you want to update.
2. Click the "..." menu at the top of the database (not the general Settings & members import).
3. Select Merge with CSV (if available for your database).
4. Choose the CSV file, then pick a matching property, a column that uniquely identifies each row (an ID, email, or SKU, not a name field).
5. Notion updates existing pages whose matching property matches a row in the file, and only creates new pages for rows that don't match anything existing.

"Merge with CSV" isn't available on every database or plan tier. If you don't see it, you'll need to clean up duplicates after a plain import instead (below).

If you already have duplicates from a previous import

Export the database back out to CSV, run it through our free Deduplicate tool to find exact duplicate rows (or Fuzzy Dedupe if the duplicates have small formatting differences like trailing whitespace or inconsistent casing), then decide which copies to keep before re-importing or manually removing the extras in Notion.

Bulk-deleting duplicates already sitting in Notion

If duplicates have already piled up across several imports and there's no reliable way to tell which copy is the "correct" one from inside Notion's UI, working from an export is faster than deleting rows by hand one at a time:

1. Open the database and use "..." → Export → CSV to pull every row, duplicates included, plus Notion's internal page ID for each one.
2. Run the export through Deduplicate, keying on whichever column should be unique (email, SKU, an ID), and keep the most recently edited copy of each duplicate set if that column is available in the export.
3. Note the Notion page IDs of the rows to remove, then delete those specific pages in Notion (select multiple rows in a table view with Shift-click, then Delete).
4. Going forward, switch to Merge with CSV (above) so the duplicate count doesn't grow again on the next import.

If you're syncing on a recurring schedule, not a one-off import

For a CSV that gets re-imported regularly (a nightly export from another system, a weekly roster update), repeating "Merge with CSV" by hand every time is fragile, someone eventually forgets, or picks the wrong matching column once and reintroduces duplicates. Notion's API supports the same update-or-create logic programmatically: query the database for a page matching your key property, and either update that page or create a new one, which is the same"merge" behavior without a manual UI step. That's a build task rather than a one-off fix, but worth knowing if this keeps recurring rather than happening once.

Doing this in How To CSV

Deduplicate and Fuzzy Dedupe both run entirely in your browser, choose the key column(s) that should be unique and get a cleaned file back with duplicates flagged or removed before it goes anywhere near Notion again.