Autonomous Underwater Vehicle Manipulator (AUVM) system which has propeller, rudder, stern, fine and manipulator as a structural component, has wide applications inside the water like exploration of the ocean resource, underwater pipe and cable inspection, geological sampling, and archaeological work. It is multivariable, nonlinear, unstable and under-actuated nature of the system and uncertainties with disturbance makes the trajectory controlling of underwater vehicle difficult. This work concentrates on the design of non-linear dynamics of AUV using the transformation matrix and Newton-Euler formulation techniques and the manipulator using Lagrange formulation with different uncertainty and extern disturbances like buoyance force and ocean current. Sliding Mode Control (SMC) were applied for this designed system for trajectory tracking of the vehicle. The high frequency oscillation (chattering) of SMC is reduced by using Artificial Fish Swarm Algorithm (AFSA) by adjusting or tuning the sliding surface parameters. The simulation was handled using Level-2 MAT-LAB s-function and Simulink to cheek the performance of the controller with the designed models. For speed control system using SMC the system tracks the desired position value with the steady state error of 0.005% in surge position and 0.0034% and 0.0266% for of joint one and joint two angular position respectively. Similarly the velocity tracks the goal velocity value with minimum error difference. Whereas with the addition of the external disturbance the AFSA reduce the chattering effect and the system tracks perfectly by the value of 0.0023% for surge position and 0.0031% for joint two angular position. For diving control the system tracks the desired value with 0.1% of steady error with disturbance and 0.01% without disturbance value of pitch position control which shows that SMC reduces the effect of the external disturbance. Mostly, the chattering effect of joint one and joint two is reduced by AFSA. Which shows that Simulink result of SMC with AFSA has better performance, more stable and robust. Key words: Autonomous underwater vehicle manipulator, Sliding Mode Controller and Artificial fish swarm algorithm.