Sorry that is meant to read 5...Nxd5 and 5...Na5
Fried Liver Attack query

That's because Nxd5 is a mistake that allows the Fried Liver and Lolli, whereas Na5 is the main line. Also ok for black is Nd4 (the Fritz variation), and b5 (the Ulvestad, which may transpose to the Fritz).
I believe the Na5 line usually continues:
5. ... Na5
6. Bb5+ c6
7. dxc6 bxc6
8. Ba4
For more information on the other lines in the Two Knight's Defense:
http://en.wikipedia.org/wiki/Two_Knights_Defense#4...d5_.28the_main_line.29

5 ... Na5 is the main line. Black gives up a pawn but gets much quicker development than White. Other playable moves are 5 ... b5 and 5 ... Nd4.
Even allowing the Fried Liver Attack (or its even more effective cousin, the Lolli Attack (6 d4)) is a serious mistake by Black, as these statistics from my roughly 5,000,000 game database show:
5 ... Na5: 6,087 games, White scores 50.7%
5 ... b5: 1,044 games, White scores 46.2%
5 ... Nd4: 998 games, White scores 49.7%
5 ... Nxd5 6 Nxf7: 1,049 games, White scores 76.7%
5 ... Nxd5 6 d4: 354 games, White scores 82.3%

The Fried Liver Attack is a draw with best play by both sides. Alas, "best play" is much harder for Black with the clock ticking. White, OTOH, can get away with a few inaccuracies. That's why the statistics are so lopsided.
Statistics do not reveal the truth of the position. Sometimes they simply offer clues to patterns of error.
I've done well on the Black side of the Fried Liver against reasonably strong opponents in correspondence chess. These games have been extremely time consuming.

That's a profound declaration. I'm not entirely convinced that we know what best play for both sides is, but I'm glad Ziryab is.
It's more a statement of faith built upon more than one hundred hours of study and practical experience. It's like Wilhelm Steinitz's declaration that chess is a draw with best play by both sides. There are plenty of opportunities to go astray, but the resources for Black in the Fried Liver should prove adequate.
My last real test in the Fried Liver started just over one year ago on this site:

omnipaul wrote:
That's because Nxd5 is a mistake that allows the Fried Liver and Lolli, whereas Na5 is the main line. Also ok for black is Nd4 (the Fritz variation), and b5 (the Ulvestad, which may transpose to the Fritz).
I believe the Na5 line usually continues:
5. ... Na5
6. Bb5+ c6
7. dxc6 bxc6
8. Ba4
For more information on the other lines in the Two Knight's Defense:
http://en.wikipedia.org/wiki/Two_Knights_Defense#4...d5_.28the_main_line.29
8 Ba4 is a mistake. 8 Be2 is the main move. Fischer messed with 8 Bd3 some 8 Qf3 is an important option.
Fair enough. I don't usually play the Two Knights anyway, preferring the King's Gambit as white and the Sicilian as black. Heck, I even prefer the Ulvestad variation those rare times I do go into the Two Knights. So I'm not too surprised there was a fault in my memory of the main line.

Yeah, 5...Na5! gives Black full comp and an easy to play position. After 6 Bb5+, Black can choose between pfren's favourite 6...Bd7, which gives Black activity and adequate compensation in a relatively simple manner, or 6...c6 7 dc bc, self destructing the Queenside pawn structure, but gaining even more time. In general, Black's lead in development and attacking prospects give him enough play, and White is usually uncomfortable in these positions unless he likes defending for whatever reason. If you don't mind facing a strong initiative (and not even being objectively better) for the sake of a pawn, 8 Bd3!? is something to look into.
I was hoping someone might be able to help me with the Fried Liver Attack. I recently played a game in which I played the attack - as such the game panned out as:
I was therefore expecting the usual 6...Nxd5 but my opponent played 6...Na5. Can someone explain why this is not the usual line for black? I know black loses a pawn but it caused me a lot more trouble than games I've played where black has played the more standard 6...Nxd5!
Thanks!
Alex