We studied the evolution of cooperation in the spatial prisoner’s dilemma game based on distributed neural networks.The simulation results showed that the individuals with neural networks could naturally develop some self-organization mechanisms,which resulted in the emergence and maintenance of cooperation.We observed that the population with small memory length developed a higher level of cooperation than that with large memory length,while the latter showed more robust cooperative behaviors than the former in the case that the noise was introduced into the system.It was also found that when the memory lengths of individuals were allowed to evolve,selection favored the medium length of memory.