BASE in Excel: Create Binary, Hexadecimal and Padded Text Codes

• Introduced: Original introduction release not established by the official history reviewed.
• Availability: official help lists Excel for Microsoft 365, Excel for Microsoft 365 for Mac, Excel 2024, Excel 2024 for Mac, Excel 2021, Excel 2021 for Mac, Excel 2019, Excel 2016.
• Checked: October 14, 2026; official BASE reference. Examples are independently calculated expectations, not native Excel test results.

=BASE(number, radix, [min_length])

number is the value being represented; radix selects the number system. The optional min_length sets a minimum output width, adding leading zeros when necessary. Omitting it produces an unpadded representation. It does not set a maximum width or truncate a longer result. Official BASE reference

The documented number domain is from zero up to, but not including, 2^53. Radix must be between 2 and 36. A supplied minimum length must be from 0 through 255. Microsoft documents truncation of fractional arguments to integers, #NUM! for constraints outside their allowed limits and #VALUE! for a nonnumeric number argument.

=BASE(45,2)
=BASE(45,8)
=BASE(45,16)
=BASE(45,36)

Expected text results: 101101, 55, 2D, 19. All four strings represent decimal 45; the underlying quantity did not change.

The hexadecimal result uses D for digit value 13: 2×16+13 = 45. The base-36 result uses two digits because 1×36+9 = 45. Keep a radix label beside exported strings so a recipient does not interpret 19 as decimal nineteen.

Enter 0, 1, 45, 128 in A2:A5. In B2 enter and fill down:

=BASE(A2,2,8)

Expected B2:B5:

These are eight-character representations for these inputs, not a promise that BASE enforces an eight-bit range. =BASE(256,2,8) is expected to return 100000000, containing nine characters.

If the destination requires exactly eight binary digits, validate the original whole number as 0 through 255 separately. Padding cannot enforce that rule.

=BASE(31.9,16.8,4.9)

Expected result: 001F. For this positive-input example, the arguments become 31, 16 and 4 before conversion. The discarded fractional portion is not rounded up.

A workflow that requires whole-number identifiers should reject fractional source data instead of relying on truncation. Otherwise distinct inputs such as 31.1 and 31.9 can produce the same text code.

Suppose A2 contains the fictional sequence number 2748. To create a readable label:

="BIN-"&BASE(A2,16,4)

Expected result: BIN-0ABC. The arithmetic check is 10×16²+11×16+12 = 2748. The prefix and padding are presentation choices; they do not encrypt the sequence or make it unpredictable.

For a second value of 4096, the result becomes BIN-1000. For 65536, it becomes BIN-10000, showing again why minimum length is not a maximum-field-width constraint.

When constructing labels, keep the numeric source in its own column. Stripping a prefix later is more fragile than retaining a clean source number alongside the display label.

For a numeric input in A2, a field accepting only whole numbers 0 through 255 can use:

=IF(AND(A2>=0,A2<=255,A2=INT(A2)),BASE(A2,16,2),"Review number")

Expected results for 15, 255, 256, 15.5 are 0F, FF, Review number, Review number. This guard is intentionally scoped to numeric inputs; it is not an all-purpose imported-text validator. AND combines the domain checks, while INT tests whole-number equality here.

BASE performs positional numeral conversion, not a signed two’s-complement encoding. Do not feed a negative number into it expecting a machine-specific bit pattern.

The result is text even when it contains only digits. Preserve that type when exporting padded codes; automatic numeric conversion can remove leading zeros. =LEN(BASE(45,2,8)) is expected to return 8, a simple output-width check using LEN.

Use the actual numeric source, not a rounded visual display, to decide whether an input is in range. A cell format can hide fractional values without removing them. Large imported numeric strings may already have lost precision before BASE sees them, so BASE cannot restore discarded digits.

Test zero, a typical value, a width boundary and a value just beyond your own business limit. The function’s broad numeric range is different from the narrower range your label format may permit.

Verification: Fictional inputs and independently calculated outputs; no native Excel execution. See the editorial testing standards.