SQL Formatting Best Practices With Examples

Most SQL bugs hide in queries nobody can read. Consistent formatting doesn't change what a query does, but it changes how quickly a human can see what it does — and whether a code review catches the mistake.

Before and after

Hard to read:

select o.id,c.name,sum(oi.qty*oi.price) total from orders o join customers c on c.id=o.customer_id join order_items oi on oi.order_id=o.id where o.status='paid' and o.created_at>='2026-01-01' group by o.id,c.name order by total desc limit 50;

Readable:

SELECT
  o.id,
  c.name,
  SUM(oi.qty * oi.price) AS total
FROM orders o
JOIN customers c
  ON c.id = o.customer_id
JOIN order_items oi
  ON oi.order_id = o.id
WHERE o.status = 'paid'
  AND o.created_at >= '2026-01-01'
GROUP BY o.id, c.name
ORDER BY total DESC
LIMIT 50;

The rules worth adopting

1. Capitalise keywords (SELECT, FROM, WHERE) and keep identifiers in lower snake_case. It makes the skeleton stand out.
2. One clause per line. SELECT, FROM, JOIN, WHERE, GROUP BY, HAVING and ORDER BY each start a line.
3. One column per line in the SELECT list once there are more than two or three. Diffs then show exactly which column changed.
4. Put each JOIN's ON condition on its own indented line. Always use explicit JOIN syntax, not comma joins.
5. Align AND / OR at the start of lines in WHERE; add parentheses when mixing them. a AND b OR c doesn't mean what many people think.
6. Alias with AS and use short but meaningful aliases (o for orders, not t1).
7. Qualify columns with the alias whenever more than one table is involved.
8. Avoid SELECT \* in application code — it breaks when columns change and fetches more than needed.
9. Use CTEs (WITH) to name the steps of a complex query instead of deeply nested subqueries.
10. Comment the why, not the what.

Trailing vs leading commas

Trailing commas are the common default. Leading commas (, column) make it easy to comment out or append a column without touching the previous line, which some teams prefer. Pick one and enforce it.

A CTE example

WITH paid_orders AS (
  SELECT id, customer_id
  FROM orders
  WHERE status = 'paid'
)
SELECT c.name, COUNT(*) AS orders
FROM paid_orders po
JOIN customers c
  ON c.id = po.customer_id
GROUP BY c.name;

Money and dates in SQL

• Store money as integers in minor units or as NUMERIC/DECIMAL, never floating point — see floating-point money bugs.
• Store timestamps in UTC and convert at the edges.
• Filter dates with half-open ranges (>= start AND < end) rather than BETWEEN.

Enforcing a style

• Use a formatter in your editor or pre-commit hook so style isn't a review topic.
• Check it in CI.
• Format queries embedded in code consistently with the rest of the repository.

Format a query instantly

Paste any query into the free SQL Formatter for clause-per-line layout, indented joins and conditions, uppercase keywords and a choice of indent and comma style — or minify it for a one-line config value. Strings, comments and quoted names are never altered, and nothing is uploaded. Compare a before-and-after with the Text Diff Checker.

When your queries outgrow the app

Complex reporting queries, slow dashboards and spreadsheet-driven data workflows are often a sign you need a proper data layer. See custom software development process and cost or talk to Pyalm.