Why does my regex match nothing at all?

Common causes: a metacharacter that should have been escaped wasn't (e.g. treating . as a literal dot), the case flag i wasn't added, or you used the anchors ^$ but the text has leading/trailing spaces or newlines. Check each one, and in a tester reduce it to the smallest matching version, then gradually add back.

Why does it match more than I want?

Usually a greedy quantifier. Change .* to the non-greedy .*?, or use a more precise character class (such as [^"]* instead of .*) to limit the range.

Regex Doesn't Match? Common Problems and Troubleshooting

Locate the Problem with a Tester First

Paste the regex and text into the Regex Tester and see what the highlighting actually matches—for the vast majority of problems, you can tell at a glance whether it’s “not matching” or “matching too much.”

Problem 1: Nothing Matches

Should have escaped but didn’t: . * + ? ( [ \ and others are metacharacters. To match a literal decimal point you must write \., otherwise . matches any character and produces the wrong logic.
Case mismatch: matching is case-sensitive by default. To match both Hello and HELLO, add the i flag.
Anchors too strict: you used ^...$ but the text has leading or trailing spaces or newlines, so the whole string doesn’t match. Try removing the anchors first, or allow whitespace.

Problem 2: Matches Too Much / Overreaches

Greedy quantifier: .* eats as much as possible. Switch to the non-greedy .*?, or use a negated character class to precisely limit it—e.g. to extract content inside quotes, use "([^"]*)" rather than "(.*)".
Forgot the word boundary: to match the standalone word cat, use \bcat\b, otherwise it will match the cat inside category.

Problem 3: Multiline Text Behaves Unexpectedly

^ $ match the start and end of the entire string by default. To make them match the start and end of each line, add the m (multiline) flag.
. does not match newlines by default. To make it match across lines, add the s (dotAll) flag.

Problem 4: Chinese / Unicode Matching Issues

\w does not include Chinese characters by default. To match Chinese, use the range [一-龥], or for certain properties use the u flag together with \p{...} (depending on language support).

Problem 5: Regex Runs Very Slowly or Even Hangs (ReDoS)

Nested quantifiers of the form (a+)+ or (.*)* cause “catastrophic backtracking” on certain inputs, with an explosive time cost.
Fix: avoid nested quantifiers, use more specific character classes, and add anchors when necessary. This site’s tester runs in a separate thread with a timeout guard, so in such cases it warns you instead of freezing the page.

Still Can’t Solve It?

Go back to basics: first use What Is a Regex to confirm what the symbols mean, follow the approach in How to Write to break the expression into the smallest fragments, validate them one by one in the tester, and then reassemble the full pattern.

Regex Doesn't Match / Matches Too Much? A Troubleshooting Guide