In this paper a model of boundedly rational decision making in the Fin
itely Repeated Prisoner's Dilemma is proposed in which: (1) each playe
r is Bayesian-rational; (2) this is common knowledge; (3) players are
constrained by limited state spaces (their Bayesian minds) in 'process
ing' (1) and (2). Under these circumstances, we show that cooperative
behavior may arise as an individually optimal response, except for the
latter part of the game. Indeed, such behavior will necessarily obtai
n in long enough games if belief systems satisfy a natural condition:
essentially, that all events consistent with the players' analysis of
the game be attributed by them positive (although arbitrarily small) s
ubjective probability.