One concern I've seen expressed with Pb free parts is the possibility of tin whiskers forming. Tin whiskers simply don't form if there's enough Pb mixed in with everything, but these days often the pins on many modern parts don't contain any Pb even if you're using lead in your solder. As far as I can tell, the industry is generally moving towards a matte tin finish for most parts.
http://nepp.nasa.gov/whisker/ has some basic background. http://www.circuitsassembly.com/pdf/0408/0408nemi.pdf seems to suggest that if you want good long term reliability wrt avoiding problems with tin whiskers, it's best to avoid parts whose lead centers are spaced less than 1mm apart. I'm not at all clear how reliable the information in this article is, though, or how often you're likely to run into problems if you choose a less conservative approach. One manufacturer's website I looked at indicated that they'd tested for tin whisker formation by keeping samples of a part in an oven at a constant temperature for some reasonably large number of hours (I think it was a bit above 1000), and tried to imply that this was a meaningful indication of whisker formation. But I thought I'd read elsewhere that thermal cycling and mechanical stresses may contribute to tin whisker growth, which makes me very uncomfortable about putting much faith into the usefulness of that manufacturer's data.
