Tracking control of robotic manipulators presents many challenges on the controller design due to the system nonlinear behavior. This paper presents the design of a tracking control for a parallel robotic manipulator using a fractional order PID controller with the feedback linearization technique. The kinematic and dynamic models of the manipulator are obtained, and a parametric identification of the dynamic model is performed using the recursive least squares algorithms. A MSC-ADAMS/MATLAB co-simulation model of the manipulator is built to perform the identification and control tasks. The inverse dynamics method gives a linearized and uncoupled model of the robotic system, which is employed to design the fractional order PID controller. The fractional order PID controller is contrasted with an integer order PID controller on performing tracking tasks. A quantitative performance analysis of the controllers is made calculating the performance indices for different tracking tasks. Obtained results show that the fractional order PID controller has a better performance on tracking tasks than the integer order PID controller.