Date Format Codes Explained

Last reviewed on 2 October 2026.

A date format code (also called a pattern, picture or format string) is a template that tells software how to write a date: DD for the day, MM for the month, YYYY for the year, and so on. The letters look universal, but they are not: mm is minutes in one tool and month in another, and YYYY can quietly mean a different year near New Year. Type any pattern into the tester below to see what it produces, then use the tables to translate it between Excel, Word, strftime, PHP, Java, .NET, JavaScript libraries and SQL. Every example on this page uses Friday 2 October 2026, 14:05:09.123.

Date format token reference

These are the building blocks used by most "human-readable" pattern styles (Excel, Moment, .NET and the generic notation found on forms). Upper/lower case matters in many tools; the cross-tool table below shows the exact letters each one expects.

Common date patterns and what they produce

The same handful of tokens combine into the patterns you see on forms, in spreadsheets and in system specifications. The patterns are written in the generic notation used on forms; the right-hand column notes where each one is typically found.

A numeric pattern on its own never tells a reader whether 02/10/2026 is 2 October or 10 February. If you control the format, prefer ISO 8601 (YYYY-MM-DD) for data and a month name (D MMMM YYYY) for people. The DD/MM vs MM/DD guide covers which countries use which order.

The same codes in Excel, strftime, Java, .NET, JavaScript and SQL

The concept is the same everywhere; the letters are not. Each cell below shows the code that produces the value in the first column.

Spreadsheets and programming languages

strftime covers C, Python (datetime.strftime), Ruby (Time#strftime), PHP's deprecated strftime() and the Unix date command. The no-padding flag %-d/%-m is a glibc/BSD extension that Python and Ruby pass through on Linux and macOS; on Windows the C runtime uses %#d instead. Excel codes are shown for English-language Excel; some localised versions use translated letters (German Excel, for example, uses TT.MM.JJJJ). In .NET, a lone d or M is read as a standard format string, so write %d when you need just the day number.

JavaScript libraries and SQL

In Oracle and PostgreSQL, Month and Day are blank-padded to a fixed width (for example October  ); the FM prefix removes the padding and leading zeros. Their case follows the code: MON gives OCT, Mon gives Oct. In MySQL, %M is the month name and %i is minutes, the opposite of what strftime users expect. In date-fns, DD and YYYY are day-of-year and week-numbering year, and are rejected unless you opt in, which is why its patterns use dd and yyyy.

Word field codes

In Microsoft Word, a date field's format comes from the \@ date-time picture switch. Press Alt+F9 to show field codes, edit the picture, then press F9 to update:

{ DATE \@ "dd/MM/yyyy" }          → 02/10/2026
{ DATE \@ "dd.MM.yyyy" }          → 02.10.2026
{ DATE \@ "d MMMM yyyy" }         → 2 October 2026
{ TIME \@ "HH:mm" }               → 14:05
{ CREATEDATE \@ "dddd, d MMMM yyyy h:mm am/pm" }

Word uses uppercase M for month and lowercase m for minutes, so \@ "dd/mm/yyyy" puts the minutes where the month should be. You can paste a whole field code into the pattern tester; it extracts the picture from \@ "…" automatically.

Specific patterns explained

Related reading

• Date converter — see one date in every common format at once.
• ISO 8601 explained — YYYY-MM-DD, week dates and why the week-based year exists.
• DD/MM/YYYY vs MM/DD/YYYY — which countries use which order.
• Intl date formatting — locale-aware formatting in JavaScript without hand-written patterns.
• Date parsing pitfalls — two-digit years, ambiguous input and lenient parsers.
• Monthly calendar — check weekdays and week numbers.