I've yet to decide how I wish to play when I'm faced with the Morra Gambit.  I usually decline it by going into the Nf6/Nc6 or Nf6/e6 branches of the Alapin (c3) Sicilian.  However, I can't help but feel that I'm letting my opponent off easy and that I should be putting a sterner test to his opening choice.  The only line of the Morra Accepted that has ever interested me goes 4...a6 5.Bc4 e6 6.Nf3 b5 7.Bb3 Bb7 with pressure on e4 and b4 soon to come. 

I think it was Nunn's Chess Openings that made a passing comment that those who are interested in looking for a refutation (of the Morra) should look at this line.  I've played it offhand in some games with success, and I like especially the idea of interrupting white's coordination by pushing the knight (or forcing a3), but I can't help feeling a bit suspicious about the whole idea: developing and loosening the queenside, leaving the kingside pieces at home, all in the name of pursuing another pawn!

I'm asking for anyone's opinion who has had any experience (on either side) with this line, or who has an idea and wants to chime in.  Also, if anyone has the newest book on the Morra (I forget the title) and could share what it says about this line, I would be very grateful. Thanks!

There's a player at my club who always plays the smith morra against the sicilian... actually he has stopped playing it against me after I continually beat him playing the line i'll suggest to you here. The line goes 1.e4 c5 2.d4 cxd4 3.c3 dxc3 4.Nxc3 e6 5.Bc4 a6 6.Nf3 Nge7 7.Qe2 Ng3  and then play your dark bishop out and caslte. There are no tricks here and your strategy is simply to exchange pieces and make the pawn up advantage a winning advantage.

Hmm...thanks for the advice.  I would very much like to see an example of the Nd5 sac with Rooks piled on the d-line, I just can't get this variation out of my head.  I think a stark visual example of Black getting thrashed would be just the right tonic for my ailment!  Would somebody indulge me?

Many thanks in advance!

p.s. I saw your correspondence game posted MNb, but maybe an example where black pushes the knight sooner.  I think the earlier the better...thanks again!  (if this doesn't work out, I think I'm back to the Alapin transposition, maybe it's less hassle).

Bangiev has written an article - I am too lazy to look for it now - on this line because of Kocsis-Bangiev, Budapest Open 1989, won by Black. This game saw 10.Rd1 b4 11.Na4 (iirc Bangiev refuted 11.Nd5) Nf6 12.Rd4 a5 and here Bangiev gives the improvement 13.Bf4 e5 (I suppose 13...Qb8 is better, but 14.Rad1 gives White enough for the pawn) 14.Bxe5 dxe5 15.Nxe5 and White is winning. So it seems Black is not that OK.
This is another gap in Palkovi's book on the Morra Gambit.

Ok, I have taken my look and in the position after 1. e4 c5 2. d4 cxd4 3. c3 dxc3 4. Nxc3 e6 5. Nf3 a6 6. Bc4 b5 7. Bb3 Bb7 8. Qe2 d6 9. 0-0 Nd7 10.Rd1 b4 11. Na4 Ngf6 12. Rd4 I think the right move is 12...Qa5.
Now white can go for the relatively quiet 13. Bd2 Qb5 14. Qe1 a5, but I fail to see a good continuation from here. It seems the most white can get is control of the open c-file after for example 15. Rac1 Be7 16. Rdc4 0-0 17. Nd4 Qe5 18. Nf3 Qh5 19. Rc7 Rab8, but the rook can be driven back by ...Bd8 and the Na4 is still out of play. Black is  better.
More in the true Morraspirit seems 13. Bf4 e5 14. Qc4 which can lead to a fantastic mess after 14...exd4 15. Qxf7 Kd8 16. Nxd4 Kc8 17. Rc1 Kb8 18. Rc6 Ra7 19. Nb6 Qxb6 20. Rxb6 Nxb6 21. Qe6. However, after the more prosaic 14...exf4 15. Qxf7 Kd8 I don't see a convincing way for white to continue the attack.

   Yes I aagree that 12.Rd4 Qa5! is better fot black so it seem that 8.0-0 instead of 8.Qe2 is the last chance

This line (without an early Bb7) was featured in a book called Smith-Morra gambit: Finegold defence by Finegold and Ciaffone. Including Bb7 only changes the position slightly.
After 1. e4 c5 2. d4 cxd4 3. c3 dxc3 4. Nxc3 e6 5. Nf3 a6 6. Bc4 b5 7. Bb3 Bb7 8. Qe2 d6 9. 0-0 Nd7 10.Rd1 b4 11. Na4 Ngf6 12. Rd4 12...Qa5 looks strong. 11.Nd5 does not work so good because Black has not played Ngf6 enabling the knight to defend the e7 square after taking on d5 opens the e-file, while Rd1 looks a bit silly.

White can possibly improve with 12.Bc2 or the speculative 12.Bf4.
At move 10 there is 10.Bf4 or 10.Nd4!? in both cases intending an improved version of the sacrifice 10...b4 11.Nd5. In the former case White also has the opstion of meeting Ngf6 with e4-e5.
I wouldn't count White out in all of these lines.

1. e4 c5 2. d4 cxd4 3. c3 dxc3 4. Nxc3 e6 5. Nf3 a6 6. Bc4 b5 7. Bb3 Bb7 and as for 8.0-0 b4? 9.Nd5!! there are long forced lines to 1-0 after accepting the sacrifice with 9..exd5 10.exd5+- Black can deviate with 9..Nc6 but after 10.Re1, white has dangerous initiative.