I think Mihail Marin was right to give the Evans a lot of respect in his book Beating the Open Games: Black has no simple way to decline the gambit and reach equality, or to accept the gambit, return the pawn and reach equality. Until the early 21st century, it was considered for a long time that 4...Bxb4 5.c3 Ba5 6.d4 exd4 7.0-0 Nge7 was the fix, but Nigel Short demonstrated in his games that 7.Qb3 is harder to crack. Of course, Black has theoretical equality, but typically White ends up with good compensation for a pawn, typically what White wants out of this opening.
In the Evans Gambit Declined, it seems that 5.b5, Captain Evans's original idea, with the idea of picking up the e5-pawn, is alright, but not a serious try for advantage, if White meets 5...Na5 with the clumsy-looking 6.Bd3. 6.Nxe5 is tactically flawed because White ends up with two pieces attacked following 6...Nh6 7.d4 (otherwise 7...Bd4 forks the knight on e5 and rook on a1) 7...d6.
Therefore White generally plays 5.a4 with the threat of trapping Black's bishop on b6 with a4-a5, and so Black usually pushes the a7-pawn to give the bishop an escape hatch on a7. My investigations of 5.a4 support the consensus view that 5...a5 is inferior because White can kick the c6-knight away from covering d4 by playing 6.b5, and then get a strong centre with a subsequent d2-d4.
5...a6 is the strongest response to 5.a4. 6.Bb2, with the idea of playing b4-b5 and not falling into tactical trouble on the a-file because the a1-rook is defended by the bishop on b2, does not appear to promise White much.
The most aggressive way to continue is 6.Nc3 intending 7.Nd5, although given that Black equalises comfortably following 6...Nf6 7.Nd5 Nxd5 8.exd5 Nd4, I suggest that White should defer the knight intrusion for another move, with 7.0-0, which makes it rather harder for Black to reach equality. Black can generally grab the e4-pawn in these lines, but if so, White gets good compensation.
The most reliable route to a slight advantage is a slow build-up reminiscent of the closed lines of the Giuoco Piano, with 6.c3 followed by d3, Nbd2 and 0-0. This is pretty risk-free and denies Black much in the way of counterplay, but may not appeal to some fans of this gambit, so I have devoted considerable coverage to both 6.Nc3 and 6.c3.
I think the biggest issue with the Evans Gambit is not the opening itself, but rather the fact that Black can avoid it with the Two Knights Defence, 3...Nf6. Most in keeping with the aggression associated with the Evans are the 4.Ng5 lines (see here and here) and the 4.d4 lines (see here and here). If any of these are suitable for you, then I can certainly recommend the Evans Gambit. There's also the argument, if it works fairly well for Nigel Short at grandmaster level, it can't be bad at club level.
Evans Declined overview
In-depth coverage (illustrative games and analysis)
