Combine Google Sheets tabs with QUERY and keep one header

Keep the first tab’s header, start every later tab at row 2, and set QUERY’s final argument to 1. Then check the combined records against each source tab. A correct-looking heading does not prove that the first transaction survived.

This workflow builds a report from monthly tabs inside one Google Sheets file. It adds reconciliation and a growth test to the QUERY function reference, rather than relying on the header setting alone.

Before you start

Use a practice spreadsheet or a copy you are allowed to edit. Create tabs named Jan, Feb, and Report. Keep Report empty. These examples use English function names and comma argument separators. Use numeric values for Month and Amount; the RecordID values are text. The checks assume the sheet has recalculated after each edit.

The four columns must mean the same thing, in the same order, on both tabs. Stacking does not match fields by their headings. Decide which field identifies a real record: here RecordID is required, while Region is optional and Amount may legitimately be zero.

1. Enter the two source tables

Enter the following cells exactly. A dash in the table means leave the cell empty, not type a dash.

Jan!A1:D6

Feb!A1:D6

Alternatively, import the January CSV and February CSV into separate practice tabs, then rename them. CSV files contain values only, without tab names or formulas. Check that the intentional empty row remains at row 4; see Google’s CSV import instructions.

2. Assemble the report with an explicit header count

In Report!A1, enter:

=QUERY({Jan!A1:D6;Feb!A2:D5},"select Col1,Col2,Col3,Col4 where Col1 is not null order by Col2,Col1 label Col1 'Record ID',Col2 'Month',Col3 'Region',Col4 'Amount'",1)

Read the formula as a small data pipeline:

1. {Jan!A1:D6;Feb!A2:D5} creates one four-column input. Only January contributes a heading row. Google’s array documentation describes vertical range joining with the semicolon inside braces.
2. The final 1 tells QUERY that one row at the top of that combined input is a header. It does not mean one header per tab. Omitting it permits inference. See Google’s QUERY header argument.
3. Col1 through Col4 address positions in the constructed array. where Col1 is not null keeps identified records. Filtering Region instead would wrongly remove J-103; filtering Amount for positive values would wrongly remove J-104.
4. order by Col2,Col1 makes the display order explicit. The label clause sets output headings; it does not rescue input rows incorrectly classified as headers. Google query-language reference.

The complete expected output is eight rows by four columns, including the header:

These are independently calculated expectations, not a captured Google Sheets result.

3. Reconcile the whole result

Use a separate Checks tab so diagnostic formulas cannot obstruct Report’s expanding output.

These count checks assume the supplied literal IDs, not formulas returning empty strings. For your real data, choose a source count that follows your record-identification rule. COUNTIF and SUMIF explain the individual checks.

Inspect J-101, J-104 and F-203 as well as the totals. All seven IDs must appear once. Matching totals alone can conceal a missing zero-value record or two offsetting mistakes.

4. Make growth part of the test

Replace the report formula with this open-ended version:

=QUERY({Jan!A1:D;Feb!A2:D},"select Col1,Col2,Col3,Col4 where Col1 is not null order by Col2,Col1 label Col1 'Record ID',Col2 'Month',Col3 'Region',Col4 'Amount'",1)

Then enter F-204, 2, North, 45 into Feb!A6:D6. The report should now contain eight data records plus its header. F-204 is the last record, February has four records totaling 225, January remains 250, and the combined amount is 475. The original bounded formula stops February at row 5, so it would omit this new record. Open-ended ranges avoid that cut-off.

Open-ended ranges can cost more calculation work. For a large file, use a deliberate shared capacity such as Jan!A1:D5000 and Feb!A2:D5000. Ensure both tabs have that capacity and review the limit before it fills. Google’s performance guidance recommends closed ranges where practical. Every additional month must be added deliberately; neither formula discovers new tabs by itself.

Diagnose failures without discarding valid records

Run each negative case separately in a disposable copy, starting with the original seven-record fixture.

For the no-input-header variant, the corrected formula is:

=QUERY({Jan!A2:D6;Feb!A2:D5},"select Col1,Col2,Col3,Col4 where Col1 is not null order by Col2,Col1 label Col1 'Record ID',Col2 'Month',Col3 'Region',Col4 'Amount'",0)

It should recover the same seven data records. If the error instead concerns array expansion, inspect the output area and move conflicting content somewhere safe. Do not erase source data to make room.

Locale, platform and testing scope

Google Sheets is a rolling service, not a numbered desktop edition. This workflow targets its web app. Formula argument separators can differ with locale; separators inside array braces have a separate role. Google documents a backslash column separator for decimal-comma locales. Keep the quoted query text intact. Review spreadsheet locale and function-language settings before adapting punctuation.

VSTACK is another way to construct the combined array, but it still needs the same column alignment, header choices and reconciliation. An Excel or other-program export is not evidence that this Google Sheets formula ran correctly.

Verification: First-party documentation checked October 16, 2026. All fixture rows, per-tab counts, totals and negative-case arithmetic were checked independently. Not tested in Google Sheets. The formulas and expected outcomes have not been verified in the application, and no application screenshots are presented. Test on a synthetic copy before applying the workflow to a live report.