Understanding the Execution Order of SQL Queries

Many SQL statements begin with SELECT , but the logical execution order of a query is different; the SELECT clause is evaluated later, typically as the fifth step. The article lists the usual order: FROM , WHERE , GROUP BY , HAVING , SELECT , ORDER BY , and LIMIT .

A diagram (referenced in the original) helps answer questions such as whether WHERE can appear after GROUP BY (it cannot), whether window function results can be filtered (they cannot because window functions belong to the SELECT clause), and when ORDER BY and LIMIT are applied (at the end).

The article notes that database engines may not follow this exact sequence; they perform optimizations that reorder operations without changing the result set.

Column aliases and GROUP BY

Some SQL dialects allow using column aliases in GROUP BY . An example is shown:

SELECT CONCAT(first_name, ' ', last_name) AS full_name, count(*)
FROM table
GROUP BY full_name

Although it appears that GROUP BY runs after SELECT , the engine can rewrite the query so that GROUP BY is evaluated first, using the original expression instead of the alias.

Optimization and reordering

In practice, engines may reorder joins, filters, and aggregations for performance. For instance, in the query:

SELECT *
FROM owners LEFT JOIN cats ON owners.id = cats.owner
WHERE cats.name = 'mr darcy'

the engine can apply the WHERE filter before performing the left join, avoiding unnecessary processing.

LINQ and pandas equivalents

LINQ queries follow a FROM … WHERE … SELECT order, illustrated with a C# example. Pandas code can be written in a similar logical sequence (JOIN → WHERE → GROUP BY → HAVING → SELECT → ORDER BY/LIMIT), though developers often reorder steps for performance.

The article also mentions that R's dplyr provides comparable abstractions for SQL‑like queries.

Original source: "SQL queries don’t start with SELECT" – https://jvns.ca/blog/2019/10/03/sql-queries-don-t-start-with-select/