I've played 3.Nc3 against the Slav for a long time.  My opinions:

1. The Schlecter is pleasant to play against.
2. You need something against 3...Nf6 4.e3 e6, which is highly theoretical and entirely fine for Black.
3. 3...dxc4 4.e4 used to score me a lot of points, but now that Vigus wrote a book on it, my opponents are prepared.  I still prefer White there, though.
4. 3...e5 is entirely playable though probably a little better for White.
5. After 3...Nf6 4.e3 Bg4, I'm not sure what is best.
6. After many years of trying to prove an edge against 3...Nf6 4.e3 Bf5 with 5.cd cd 6.Qb3, I've basically concluded that Black is fully equal after 6...Bc8!  The trouble for White is that he has played 4.e3, blocking in his Queen's Bishop, his Knight is on c3 so you can't play the Bc1-d2-b4 plan, and the Queen on b3 is now misplaced.  After something like 7.Nf3 e6 8.Bd3 Nc6, Black will later play ...Nc6-a5, regaining the tempo he lost earlier; the Queen doesn't really have a better square than d1 to move to.  If White tries for IQP positions with the e3-e4 push at some point, which looks like a logical way to exploit Black's wasting two tempos, the Queen is still misplaced on b3 and White has lost a move playing e2-e3 and then e3-e4.  So again, Black is fine.

Just my two cents, and why I head for the Dutch variation these days.

Alias wrote on 08/19/11 at 13:47:05:

Which, at the risk of stating the obvious, transposes to the Meran Semi-Slav after 5.Nf3, so you'll need to study that or deviate at some point.

Vigus - Play The Slav http://www.everymanchess.com/chess/books/Play_the_Slav

He advocates 3...dxc4 vs 3.Nc3. Several older books acknowledge the problem of 3...Nf6 4.e3 Bf5 and suggests 4...e6. The very recent Lakdawala book seems to suggest 4...a6 judging from the intro, although there is a section on 3.Nc3 dxc4.

a) Not so easy to achieve +=, see Vigus.
b) Not trendy.
c) Don't count on 4...Bf5. It's a well known inaccuracy. 4...e6, 4...a6 and 4...g6 are more likely.