I have a file where the heading line is duplicated within the file, for example: HEADING 11111 2222222 HEADING 333 I need to delete all occurrences of the heading line after the first one, so that I am left with HEADING 11111 2222222 333 I can't seem to get it in sed. In vi/vim/ed/ex it would be 2,$g/^HEADING/d i.e. from line 2 to the end of the file, find occurrences of ^HEADING and delete them. However this syntax is not valid in sed: $ sed '2,$g/^HEADING/d' testit.dat sed: -e expression #1, char 5: Extra characters after command Any ideas? It doesn't have to be sed, but the file is almost 4GB so I think vi and ed are out unless there is an option I'm missing. Thanks

It's not beautiful, but it works: sed -e '2,$s/^HEADING$/non-existing-string/' -e '/^non-existing-string$/d' testit.dat Luke Chi

On line #1, just branch to the end. sed -e 1b -e /^HEADING/d testit.dat

Thanks James, never knew about the b command. Neat hack, Luke ;)