The book "Opening for White according to Kramnik" Vol. 3 by Khalifman could provide an answer for us on this variation, but unfortunately I do not have it. Anyway, I agree Björn that 8...a5 is strategically richer than 8...Na6 9.Rb1 Nc5!? 10.b4 Nce4. Thanks for the reply.

In the main line of the dutch leningrad

1.d4 f5 2.c4 Nf6 3.Nc3 d6 4.g3 g6 5.Bg2 Bg7 6.Nf3 0-0 7.0-0 (or in any other move-order)

7...Qe8 8.d5 is a major continuation. 

Now black must decide between the traditional 8...Na6 and 8...a5. The latter seems to have become increasingly popular to avoid problems in the 8...Na6 9.Rb1 lines, but against most other 9th moves from white, I still prefer 8...Na6 (and therefore try to keep it) because the move a5 is often not really necessary.

My question is:
After 7...Qe8 8.d5 Na6, Pedersen's recipe (in his book "The Dutch for the attacking player", 1996) against the critical 9.Rb1 was 9...Nc5!? 10.b4 Nce4. Isn't it just a good way to solve the problems of 8...Na6, or is this variation having some problems too? I suppose something must have come up recently, otherwise Kindermann gm, an expert, would not have felt it necessary to leave 8...Na6 and switch to 8...a5 (which was a difficult decision - as he writes in his book "Leningrader System", 2002). 

I hope someone has an answer. Thanks a lot in advance.

feb