Back up / test first. This deletes data. Run the inner SELECT on its own to see exactly which rows will be removed before you run the DELETE.

Number duplicates with ROW_NUMBER and delete everything past the first. This works in MySQL 8+, PostgreSQL, and DuckDB (the table needs a unique row id):

DELETE FROM users
WHERE id IN (
  SELECT id FROM (
    SELECT id, ROW_NUMBER() OVER (PARTITION BY email ORDER BY id) AS rn
    FROM users
  ) t
  WHERE rn > 1
);

-- SQLite: keep the lowest built-in rowid per key
DELETE FROM users
WHERE rowid NOT IN (SELECT MIN(rowid) FROM users GROUP BY email);

-- PostgreSQL: shortcut using the system ctid (no id column needed)
DELETE FROM users a
USING users b
WHERE a.email = b.email AND a.ctid > b.ctid;

-- BigQuery / Snowflake: rewrite the table with only the keepers
CREATE OR REPLACE TABLE users AS
SELECT * EXCEPT (rn) FROM (
  SELECT *, ROW_NUMBER() OVER (PARTITION BY email ORDER BY id) AS rn
  FROM users
)
WHERE rn = 1;

The MySQL “target table” trap. MySQL refuses to DELETE from a table you also read in a subquery — “You can’t specify target table ‘users’ for update in FROM clause”. The double subquery above (SELECT id FROM (SELECT … ) t) forces MySQL to materialize the inner result first, which sidesteps it. (SELECT * is fine in BigQuery/Snowflake but BigQuery has no ctid/rowid, hence the rewrite form.)

Delete duplicate rows, keeping one