Your regex is fine, though it could be said more economically. The problem is that you don't assign the results of a match to a list or array, my @nums = $split[1] =~ /^(\d{10})(\d{10})(\d{10})$/; Your substitution is throwing away the second two numbers.

Since you have fixed-width fields and no seperators, unpack seems natural. my @nums = unpack 'a10 a10 a10', $split[1]; You know the format of the fields, so you really shouldn't have to worry about whether they are digits (Assuming you don't have to detect bogus data).