**Performance:**

- EXISTS often faster than IN for subqueries

- Correlated subqueries evaluated per row (slow)

- CTEs may be materialized (good or bad)

- Consider JOINs for better optimization

- Test query plans for large datasets

EXISTS often faster than IN

Avoid correlated subqueries when possible

Consider JOINs for large datasets

Test CTE materialization behavior

Use EXPLAIN to analyze query plans

Correlated subqueries can be very slow

IN with NULL in subquery returns no rows

Forgetting to filter NULL in subqueries

Overly complex nested subqueries

Not testing performance with production data