JOIN in Google Sheets: Combine Lists With a Delimiter

=JOIN(delimiter, value_or_array1, [value_or_array2, ...])

The delimiter specifies the separator. The first value or one-dimensional array is required; additional values or arrays are optional. Google’s JOIN reference describes row or column arrays, rather than an unrestricted two-dimensional table contract.

Use "" to make the no-separator choice explicit. Google also documents an empty delimiter argument slot. The examples here use quoted separators for readability, including spaces within the delimiter when needed.

Enter this table in A1:C4:

In D1 enter Display code. In D2 enter and fill down:

=JOIN("-",A2:C2)

Expected results are North-Blue-Large, South-Green-Small and West-Red-Medium. Each input range is a single row. The hyphens come from the delimiter, so they do not need to be stored in the source cells.

To summarize the region column in one cell:

=JOIN(", ",A2:A4)

Expected result: North, South, West. This time the input is one column. The comma and following space are both part of the separator.

With the same table:

=JOIN(" | ",A2:A4,"End")

Expected result: North | South | West | End. The extra scalar is appended after the column values. A delimiter is inserted between adjacent elements; it is not an automatic prefix or suffix.

For a no-separator example, =JOIN("","AB","12","Z") gives AB12Z. That may be convenient for a display identifier, but it is not necessarily a unique key: joining "A","BC" and "AB","C" without separators produces the same text ABC. Even a separator is insufficient if it can occur ambiguously within the components. Design identifiers around the data’s actual constraints.

For the first display row, use a multi-character separator:

=JOIN(" :: ",A2:C2)

Expected result: North :: Blue :: Large. If that result is in F2, this formula is intended to recover the three fields:

=SPLIT(F2," :: ",FALSE,FALSE)

Expected output across three cells: North, Blue, Large. Google’s SPLIT reference documents FALSE for treating the delimiter as a whole string and the second FALSE for retaining empty fragments.

This is a controlled demonstration, not a general reversible serialization method. If a component itself contains ::, the reconstructed fields would be ambiguous. JOIN does not by itself supply CSV quoting, escaping or a data interchange schema.

JOIN has no ignore_empty parameter. Do not add a TRUE argument expecting it to control blank handling; it occupies a value position instead.

If skipping empty cells is an explicit requirement, TEXTJOIN provides that parameter. For example, with North, a genuinely empty cell, and Large in H2:H4, =TEXTJOIN("-",TRUE,H2:H4) has documented expected result North-Large. The JOIN page does not provide a complete distinction among empty cells, empty strings and other coercions; those edge cases were not natively verified here.

Check range shape first. Process a single row or column instead of assuming a rectangular block will flatten in a particular order. Then inspect the delimiter, source whitespace, text quoting and locale argument separators. Include any intended spaces inside the delimiter itself.

Numbers and dates needing a particular display should be converted deliberately with TEXT before joining. The guide does not promise preservation of source cell formatting. A full native error-token matrix and array expansion behavior were not established.

JOIN is documented for the rolling Google Sheets service. Its original introduction date, first release and last functioning numbered version were not established. Similar names in other applications do not establish the same array or blank-handling contract.

NOT_TESTED_IN_GOOGLE_SHEETS. Individual Google references for JOIN, SPLIT, TEXTJOIN and TEXT were inspected on October 17, 2026. Original string assembly and the controlled split were independently checked; no native execution or screenshot is claimed.

Related published guides: SPLIT, TEXTJOIN and CONCATENATE.