Syntax Updated November 2026

Negated character class

[^...] matches one character that is not in the set: [^0-9] is any non-digit, [^aeiou] any character except a lowercase vowel. The ^ negates only when it is the first character inside the brackets. Negated classes match newlines too, which makes [^>]* a reliable way to read up to a delimiter.

Syntax

[^set]

Examples

"a1b2c3".replace(/[^0-9]/g, "")

Result: "123"

Remove every character that is not a digit.

"sky myth dry hello".match(/\b[^aeiou\s]+\b/g)

Result: ["sky", "myth", "dry"]

Words with no lowercase vowel; \s is excluded so the match stays inside one word.

'key="value" x'.match(/"([^"]*)"/)[1]

Result: "value"

"a=1;b=2;c=3".match(/[^;]+/g)

Result: ["a=1", "b=2", "c=3"]

Common mistakes

Placing the caret later, as in [a^b], where it is a literal caret.
Expecting [^x] at the end of the string to succeed: it still needs one character to consume; use (?!x) for a zero-width check.
Using [^abc] to mean "not the word abc"; it excludes single characters. Use a negative lookahead for words.
Forgetting that a negated class also matches newlines, which can let a match run onto the next line; add \n to the set when that matters.

Practice Negated character class

These drills check your answer on live data, free, no account needed.

Negated class: no vowelsThe whole line must match. Match lines that contain none of the lowercase vowels a, e, i, o or u. Any other characters are fine, including uppercase letters, digits and spaces.
Matching quote marksThe whole line must match. Match text wrapped in quotes, where the closing quote is the same kind as the opening one: both double or both single.
key=value pairsLines hold key=value pairs separated by semicolons. Capture every pair: the key (word characters) in group 1 and the value (everything up to the next semicolon, possibly empty) in group 2.

Questions

How do I match everything except a specific character?

[^x]+ matches a run of characters that are not x. For a set, list them all: [^,;]+ stops at a comma or semicolon.

How do I match everything except a word?

A class cannot do it; use a negative lookahead: ^(?!.*\bword\b).*$ matches lines that do not contain the word.

Does [^x] match a newline?

Yes, which is why [^"]* is handy for quoted strings that span lines. Write [^"\n]* to stop at a line break.

Character classesWhat a character class matches, how ranges and negation work, what \d \w \s and their uppercase opposites mean, and what the dot does and does not match.
Lookahead (?=...) and (?!...)How lookahead checks what follows without consuming it: numbers before a unit, words without a prefix, thousands separators and file name filters.
Lazy vs greedyHow greedy quantifiers overshoot, how adding ? makes them lazy, why a negated class is often better than .*?, and the performance and newline traps.
Regex for HTML tagsHow to remove HTML tags with <[^>]+>, list tag names, read an href value, match the content of one tag lazily, and why regex is not an HTML parser.