An optimal decision making strategy can be difficult to identify in a dynamic environment, because the outcome of a given action may be stochastic and temporally delayed. Therefore, to estimate reliably the desirability of a given action, signals related to one’s choice and its outcome must be integrated across time. Although reinforcement learning algorithms provides efficient solutions to these problems, the corresponding neural mechanisms are still largely unknown. Here, we show that the primate prefrontal cortex might be involved in the temporal integration of signals related to the animal’s choices, suggesting that they might provide eligibility trace necessary to resolve the problem of temporal credit assignment. In addition, many neurons in the prefrontal cortex dynamically modulated their activity according to the outcome of the animal’s previous choice. These results suggest that the primate prefrontal cortex can play a key role in evaluating the relative benefits of choices, and thereby enabling the animal to optimize its decision making strategy.