A terminal sliding mode control method with adaptive fuzzy system was proposed for the robotic manipulator trajectory tracking control problem with a large amount of uncertain information such as modeling errors and external disturbances. The nonsingular terminal sliding surface was adopted to ensure global fast convergence of state variables during sliding stage in this method. An improved double exponential reaching law with variable coefficients was used to improve the convergence rates of state variables and suppress the chattering of controller output during the approaching stage. An adaptive fuzzy multiple-input multiple-output (MIMO) system was utilized to approximate the system model and external disturbance, in order to get rid of the dependence on model information and improve trajectory tracking accuracy as well as anti-disturbance performance. The closed-loop stability and finite-time convergence of the system were proved by constructing Lyapunov functions. The Denso VP6242G serial manipulator was taken as the controlled object for comparative simulation and experiment. Results showed that the designed controller can effectively improve the trajectory tracking accuracy and anti-disturbance ability, and alleviate the chattering phenomenon in the controller output as well.