Simulation studies in the context of Robert Axelrod’s research on iterative prisoner’s dilemma games focus nearly exclusively on the two-player- version of the game. In contrast, this article reports results of a simulation with an iterated N-prisoners’ dilemma where group size N varies between 2 and 30. The simulation investigates the relative performance of conditional cooperative strategies with increasing group size. Results show that some ‘nice’ strategies like ‘tit-for-tat’ are relatively successful and robust even in larger groups and non-nice environments. (...) However, this does not solve the cooperation problem. On the contrary, the relative success of some ‘nice’ conditional cooperative strategies is paralleled by a rapid decline of cooperation in large groups. (shrink)