Robots are highly nonlinear and chaotic in position control. The present paper mainly presents the position control of PUMA-560 Robot manipulator. Computed torque controller (CTC) is one of the solutions for position control of robot manipulators. The main drawback of controller is that it fails to operate under different dynamic operating conditions. To overcome this difficulty, intelligent controllers have gained importance. In this paper a novel approach, design of prisoner's dilemma-based fuzzy C-means controller to control the position of robot manipulator is presented. This controller is employed at the inputs of computed torques for obtaining the desired position. Fuzzy C-means controller with computed torques is realised by validating the clusters to choose most contributed rules. Thus the unfired rules are eliminated from the actual rule-base. Hence, a compact fuzzy controller with minimum rule-base, fuzzy C-means computed torque controller (FCMCTC), is designed. The concept of prisoner's dilemma is introduced in this paper to improve the fuzzy strategy. The interrelations between inputs and outputs of a fuzzy linguistic model are assigned using payoff matrix through prisoner's dilemma. Simulation results prove the efficacy of proposed controller when compared to proportional derivative computed torque controller (PD-CTC), normal FLC and that of the reference signal.