Back to the description of the bubble sort that Wirth had in his book (the book is "algorighms + data structures = programs," for the curious), he suggests modifications to a bubble sort: Always remember the position of the last switch that you've made. You can start/end here next time. (A similar thing was mentioned in a previous mail by somebody) Alternate the order of the sort going right-to-left one pass and left-to-right the next. That takes advantage of the fact that a number that's incredibly far out of place will tend to move further if you're sorting towards its proper place. (In other words, it's as you say, but assuming that we don't know the order of the data, so the sort can be generic and still be more efficient) Regards, Chris Christopher Smith, Perl Developer Amdocs - Champaign, IL /usr/bin/perl -e ' print((~"\x95\xc4\xe3"^"Just Another Perl Hacker.")."\x08!\n"); '