I know two pawnless endgames that go over the 50-move rule.
King and 2 Bishops vs. King and Knight = 70 moves average.
King, Rook and Bishop vs. King and 2 Knights = 200 move average.
What others are there?
Chess Stack Exchange is a question and answer site for serious players and enthusiasts of chess. It only takes a minute to sign up.Sign up to join this community
According to wikipedia http://en.wikipedia.org/wiki/Pawnless_chess_endgame, section "Endgames requiring more than 100 moves to win", the longest is
R+N vs N+N = 243 moves with 78% of winning chances.
R+B vs N+N = 223 moves with 96% of winning chances.
R+N vs B+N = 190 moves with 72% winning chances.
and the list goes on...
Another way is looking at the database of mutual zugzwangs at http://chess.jaet.org/cgi-bin/mzugs?sort=wmaxdtm sorted by maxdtm (depth to conversion or mate).
John Nunn's book "Secrets of pawnless endgames" (I love that book) in the 2nd editions expands coverage to some of the endings mentioned (although very briefly).
If you mean "hardest" == "longest": There are 7-man table bases available and it is said that the longest pawnless ending takes 545 moves. It is KQN vs. KRBN.
And in case you are too lazy to go through the moves by yourself :-)
I think that is what is known today. I would not be surprised if even longer pawnless endings exists with 8 pieces on the board.