Remove Lines Containing a Word or Text

How to remove lines containing text

1. Paste your text into Your text, or use Open file for a .txt, .csv or .log file. Example loads a short mailing list.
2. Type what you are looking for in Text to look for. For several words, separate them with commas.
3. Under Lines that match, choose Remove them to delete the matching lines, or Keep only them to delete everything else.
4. Tick Case-sensitive, Whole line must match or Regular expression if you need a stricter match.
5. The result updates as you type. The status line shows how many lines matched; Copy, Download .txt or Use as input to run another tool on it.

What each option changes

Whole line must match works with both kinds of query. With plain text, each comma-separated term is compared with the whole line, so a, b matches the lines a and b (write a\, b for a line that reads a, b). With Regular expression, the pattern must match the entire line, like grep -x: cat matches cat but not catalog. In regex mode spaces at the ends of the line are not ignored, so cat does not match Cat ; write \s*cat\s* if they may be there.

Exactly how lines are matched

• A line is everything between two line breaks. Windows (CRLF), old Mac (CR) and Unix (LF) breaks are all recognized, and the result always uses LF.
• Contains means the text appears anywhere in the line, inside other words too: cat matches concatenate. Use the regex \bcat\b for the word on its own.
• Several terms: unsubscribe, bounced is two terms and a line is matched if it contains either one. Spaces next to a comma are ignored; all other spaces are part of the search, also at the start or end of the query: bar, zzz searches for bar (with a leading space, so it matches foo bar but not foobar) and zzz.
• Commas: write \, to search for a literal comma inside a term (Smith\, John). A query made only of commas, such as ,, is searched as typed and matches every line that contains a comma.
• Empty lines never contain a word, so in Remove mode they stay, and in Keep mode they go. To remove blank lines, use the regex ^\s*$ or the remove empty lines tool.
• Order and duplicates: the remaining lines keep their original order, and identical lines are all kept or all removed. Run the duplicate line remover afterwards if you need each line once.
• Accents and emoji are compared as typed: cafe does not match café. Emoji can be searched like any other character. When case is ignored, the Turkish capital İ counts as i, so istanbul matches İstanbul.
• An empty query removes no lines.
• An invalid regular expression, such as an unclosed (, shows a message like “Invalid regular expression: unterminated group” instead of a result. Escapes that grep and Notepad++ accept but JavaScript’s Unicode mode rejects, such as \:, \-, \# or \@, are read as the plain character.

Worked examples

The built-in example, with Text to look for set to unsubscribe, bounced and Remove them:

Before                              After
[email protected]                    [email protected]
bounced: [email protected]            [email protected]
[email protected]                     [email protected]
unsubscribe: [email protected]
[email protected]

A server log with the query debug, trace. Case does not matter, so both DEBUG and debug go:

2026-10-21 09:00:01 INFO  Server started
2026-10-21 09:00:02 DEBUG Loading config
2026-10-21 09:00:05 WARN  Disk 85% full
2026-10-21 09:00:07 debug Cache warm
2026-10-21 09:01:10 ERROR Payment failed

becomes

2026-10-21 09:00:01 INFO  Server started
2026-10-21 09:00:05 WARN  Disk 85% full
2026-10-21 09:01:10 ERROR Payment failed

With Case-sensitive ticked and the query DEBUG, only the 09:00:02 line is removed and the lowercase debug line stays. With Regular expression, WARN|ERROR and Keep only them, the result is just the WARN and ERROR lines.

Plain text and a regex give different results for keyword lists:

With Whole line must match and the query n/a, the lines N/A and n/a (with trailing spaces) are removed, while N/A pending stays because it has more text.

When people need it

• Cleaning a mailing list: remove addresses marked bounced or unsubscribe in an export, or keep only one domain with @example.com and Keep mode. To pull addresses out of running text first, use the email extractor.
• Reading logs: drop DEBUG and health-check lines, or keep only WARN and ERROR lines.
• SEO keyword lists: remove negative keywords such as free, jobs or a competitor’s brand from a keyword export, then sort the lines.
• URL lists: remove tag pages and tracking links from a crawl with /tag/, utm_.
• Config and code: remove comment lines with the regex ^\s*# (lines starting with #, indented or not), while lines such as code # x stay.

To change text inside the lines instead of deleting whole lines, use find and replace.

The same job in grep and Notepad++

On Linux and macOS, grep -v 'text' file.txt prints the lines that do not match, which is this tool’s Remove mode; plain grep is Keep mode. -i ignores case (this tool’s default), -x matches the whole line and -F treats the pattern as plain text. Note that grep is case-sensitive unless you add -i, while this tool is case-insensitive unless you tick Case-sensitive.

In Notepad++, open Search > Mark, type the text, tick Bookmark line and click Mark All. Then choose Search > Bookmark > Remove Bookmarked Lines to delete those lines, or Remove Non-Bookmarked Lines to keep only them. This page needs no install and gives the same result for text copied from any program.

Questions