This chart is from Andrew Gelman's fascinating Red State, Blue State blog, the book of which I will get around to reading this winter break. The numbers mostly speak for themselves, but I do have a couple of notes:
1) The chart doesn't necessarily imply that all young voters voted Obama--they could have voted Libertarian or Green or even have been more likely to write themselves in. But given the totals that third parties garnered overall, most of those non-Republican votes were probably for Obama.
2) Are young voters more likely to be contrarian in general? In each of these elections, voters aged 25 have been more likely than voters aged 65 to vote for the non-incumbent party. Although the effects are obviously bigger in 2004 and 2008, the fact that this holds in 2000 as well for the opposite party poses some questions.