QUERY uses the Google Visualization query language, with explicit assumptions about headers and column types. It is not a general SQL engine accepting every database query unchanged.
State the input and header count
For a table in A1:C10 with one header row, =QUERY(A1:C10,"select A, C where B = 'Open'",1) selects columns A and C from qualifying rows.
The final 1 identifies one input header row. If omitted or set to -1, Sheets guesses, which can misclassify a short dataset. Use 0 when the selected input genuinely has no header.
Check column identifiers and types
QUERY supports column-letter and Col notation in appropriate data contexts. Make sure the identifiers correspond to the range or array you actually supplied, not a visible label typed in the header.
Each query column has a type. In a mixed-type column, the majority type determines interpretation and minority types become null for query purposes. Formatting numeric-looking text does not necessarily make the input uniformly numeric.
Test the selected population
Begin with a simple select before adding grouping, ordering, or pivots. Compare one included row, one excluded row, and the output count with the original dataset.
Inspect mixed types and blanks before accepting totals. Keep source data unchanged and use helper columns when normalization is needed. Read the query-language syntax for dates and aggregation rather than borrowing an unsupported SQL expression. A query that runs without an error can still summarize the wrong column or lose minority-type values, so reconcile the result with a small independently checked sample.
References
Reviewed against the linked official product guidance on November 3, 2026. Features, limits, screen labels, and access can vary by account, platform, region, and organization policy.