Advise and consent
Fabio Rojas resumes his series of posts of advice for grad students with one on choosing an advisor.
Piled Higher and Deeper cautions against doing too much of the kind of homework Fabio recommends. I'd err on the side of too much not too little homework, here; I think this is a rare case in which PhD gets things noticeably wrong (though in the service of a funny and true point about advisee-advisor life). As soon as a grad student sees an advisor's CV, which is a pretty routine occurence, he or she would fall into the "google-stalking" category-- that can't be right.