I do listen to music sometimes when programming, but it's interesting to note the study cited in DeMarco and Lister's Peopleware (IIRC) where they found program

I do listen to music sometimes when programming, but it's interesting to note the study cited in DeMarco and Lister's Peopleware (IIRC) where they found programmers (in a ridiculously contrived study) who listened to music produced "less inventive" solutions than those who didn't. I've always taken that as an interesting caveat.