My investigations into the Evans Gambit Accepted are concluded with a look at the most critical lines following 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.b4 Bxb4 5.c3 Ba5 6.d4 exd4. The illustrative examples are now complete as well.
After this I envisage taking a little break from the Evans and looking at the Cochrane Gambit and maybe one or two other "lighter" lines before returning to the Evans and covering the declined variations, as 4...Bb6 in particular will quite often be encountered. I also think it will be well worth revisiting my coverage of the King's Gambit, having read most of John Shaw's book and seen some discussion of the relevant lines at Chesspublishing.com.
The main problem with 7.0-0 has always been 7...Nge7 preparing to strike out in the centre with ...d7-d5. In these lines White can generally regain the gambit pawn but Black has little trouble equalising.
For instance 8.cxd4 d5 leaves White with nothing better than 9.exd5 Nxd5, and White can typically regain the pawn by playing Qb3 and Qxb7 but Black ends up quite comfortable.
More aggressive is 8.Ng5 intending 8...d5 9.exd5 Ne5 (not 9...Nxd5? 10.Nxf7) and then White should play 10.Qxd4 rather than the retreat 10.Bb3. Play can get quite complicated, but the complications quite often burn out to a draw or an equal endgame.
Thus Nigel Short has preferred 7.Qb3, attacking f7 and preventing Black from carrying out the ...Nge7 and ...d5 plan immediately. Play then typically continues 7...Qf6 8.d4 Bb6 9.e5 Qg6 10.cxd4.
This offers the sacrifice of a second pawn on d4, which gives White dangerous compensation if accepted, so Black generally acquires the bishop-pair with 10...Na5, but 11.Qa4 Nxc4 12.Qxc4 still gives White reasonable compensation for the pawn due to the strong centre and open lines for the pieces. Mihail Marin recommended 8...Nge7 but I am not sure that Black's life is so easy after 9.cxd4 Bb6 10.e5 Qf5 11.Ba3. Black can try to return the pawn with ...d7-d6, but it is not a definite equaliser.
I think 7...Qe7 is also similarly playable, leading to dynamically equal play, but it is perhaps harder for Black to handle the resulting positions over the board. 8.d4 Bb6 (8...d6 and 8...Nf6 currently look likely to concede some advantage to White, in spite of the pawn minus) 9.cxd4 Nxd4 10.Nxd4 Bxd4 11.Nc3 is the usual continuation which gives White dangerous compensation for two pawns.
In conclusion, Black appears able to keep the chances level after 4...Bxb4 5.c3 Be7, 5...Bc5 and 5...Ba5 but in each case White appears able to ensure that it is a dynamic rather than sterile type of equality, so the accepted lines of the Evans are looking in reasonable shape from White's point of view as well. I would say that 7.Qb3 is looking more promising after 5...Ba5 6.d4 exd4 because although the resulting positions are only equal, it is much harder for Black to return the pawn and reach easy equality. I think that 5...Bd6 is interesting, but with best play White should be able to get some advantage against it.