DELETE FROM in PostgreSQL

DELETE FROM removes rows from a table.

Basic Syntax

DELETE FROM table_name WHERE condition;

Example

DELETE FROM employees WHERE id = 5;

Just like UPDATE, omitting WHERE deletes every row in the table:

DELETE FROM employees;
-- removes ALL rows — the table itself still exists, just empty

If you genuinely need to empty a large table, TRUNCATE TABLE employees; does the same thing far faster (see the DROP vs TRUNCATE lesson).

Deleting Based on Another Table

DELETE FROM orders o
USING customers c
WHERE o.customer_id = c.id
AND c.is_banned = true;

RETURNING the Deleted Rows

DELETE FROM employees
WHERE department = 'Support'
RETURNING id, name;

Common Errors & Troubleshooting

• Forgetting WHERE and deleting every row — the same trap as UPDATE; run the equivalent SELECT first to confirm which rows match before deleting.
• ERROR: update or delete on table "customers" violates foreign key constraint — you tried to delete a row that other tables still reference; delete/reassign the dependent rows first, or set ON DELETE CASCADE/SET NULL on the foreign key if that's the intended behavior (see the Constraints lesson).
• DELETE FROM orders o USING customers c WHERE ... deleting more than expected — double-check the join condition in the USING/WHERE clause is selective enough; a loose join condition can match (and delete) far more rows than intended.

Key Takeaways

• DELETE FROM table; with no WHERE removes every row — always confirm the condition first.
• For emptying an entire large table, TRUNCATE is faster than an unconditional DELETE.
• RETURNING lets you see exactly which rows were deleted, in the same statement.