Barry Leiba (a fellow Indy Science Blogger) has an interesting discussion on whether it matters where you go to college. If you are a good student, does the success of your future career really depend on where you go?
Barry makes some good points about this issue from the persepctive of someone working as a computer scientist for a tech company. I think much of what he says applies to academia as well - if you browse through the faculty pages of top university science departments, you'll see most people in those departments got their PhDs in top schools. I think a good student from just about any serious 4-year college could get into a good graduate science program, but where you do your graduate work has a big impact on your career prosepcts.
For another take on this issue, looking at law schools, check out Balkanization.
One thing I hate is top ten lists. Barry links to one he takes issue with. I take issue with the idea that ten is some magical number - that ten schools are somehow in a different league from the rest. I think a better number would be twenty - there are so many very good universities in the US and Canada, and a prosepctive student should take a broader look at what's out there.