Formula
Remove the column that should not define uniqueness before calling UNIQUE:
=UNIQUE(CHOOSECOLS(A2:C6, 1, 3))
This example compares columns 1 and 3 and ignores column 2.
Test setup
We entered five synthetic rows with an ID, a timestamp to ignore, and a status:
| ID | Ignored timestamp | Status |
|---|---|---|
| A-100 | 2026-07-01 09:00 | Open |
| A-100 | 2026-07-02 10:00 | Open |
| B-200 | 2026-07-03 11:00 | Closed |
| B-200 | 2026-07-04 12:00 | Closed |
| C-300 | 2026-07-05 13:00 | Open |
The formula was run twice in separate, non-overlapping output areas.
Tested steps
- Identify the columns that should define a unique row.
- Use
CHOOSECOLSto create an array containing only those columns. - Wrap that result in
UNIQUE. - Leave enough empty cells for the result to expand.
- Repeat the formula in a separate empty area.
Both live runs returned the same three rows: A-100 / Open, B-200 / Closed, and C-300 / Open.
Why not pass the full range to UNIQUE?
UNIQUE compares the rows in the range it receives. If a changing timestamp remains in that range, otherwise identical rows are still different. CHOOSECOLS creates a smaller array first, so the ignored timestamp never participates in the comparison.
Spill-range warning
The first attempt deliberately exposed a useful failure mode: placing the second formula inside the first result’s expansion area caused #REF!. Move the formulas apart or clear the blocked cells; do not treat that error as a deduplication failure.