How do I remove duplicate result rows with SELECT DISTINCT?

Use DISTINCT after SELECT, and select only the column whose repeated values you want collapsed. For example, suppose this invented sample data is in a SQLite table named events:

id | event_label | venue
1  | Workshop    | Hall A
2  | Workshop    | Hall A
3  | Workshop    | Hall B
4  | Reading     | Hall A

To return one result row per distinct label, write:

SELECT DISTINCT event_label
FROM events;

The result is:

event_label
-----------
Workshop
Reading

The database compares the selected result rows when applying DISTINCT; here each result row contains only event_label. [S1] This removes duplicate rows from this query’s output. It does not delete or change any records in events.

Add venue to the selected columns, but note that this changes what counts as a duplicate. SQLite keeps a row for each distinct combination of selected values. So the two Workshop rows at Hall A collapse into one, while Workshop at Hall B remains separate. [S1]

SELECT DISTINCT event_label, venue
FROM events;

With the sample data above, the result is:

event_label | venue
------------|------
Workshop    | Hall A
Workshop    | Hall B
Reading     | Hall A

This query does not mean “one row per event label, with an arbitrary venue.” It means one row per unique event_label and venue pair. If the goal is one row per label, select only the label; if the goal is one row per label-and-venue pair, select both.

Start by describing what one output row should represent. Then select the columns that define that row. For instance, “one row per label” maps to SELECT DISTINCT event_label; “one row per label at each venue” maps to SELECT DISTINCT event_label, venue.

Before adding a display column, ask whether two records with different values in that column should still count as the same output row. If yes, adding it to this SELECT DISTINCT can keep multiple rows where you expected one. If you need both one row per label and some additional information, clarify how that information should be chosen or summarized; DISTINCT alone does not specify a representative venue or a rule for selecting one.

Yes. Replace the example table and column names, then check that the selected columns match your definition of a duplicate. SQLite’s documented DISTINCT comparison rules treat two NULL values as equal; text comparison uses the applicable collation. [S1] The source explains SQLite behavior, not every SQL database, and it does not establish which query is appropriate for your data or guarantee how another database handles comparisons. Check your database’s documentation if you are using a different system.

Copyable pattern:

-- One output row per label
SELECT DISTINCT event_label
FROM events;

-- One output row per label-and-venue pair
SELECT DISTINCT event_label, venue
FROM events;

In the second query, every selected value participates in deciding whether a result row is a duplicate.