This paper presents the development of remotely operated vehicle (ROV) control modelling and control synthesis using nonlinear adaptive U-model and compares it with the proportional-integralderivative (PID) control and fuzzy logic control (FLC) approaches. A nonlinear ROV model based on dynamic equations using the Newtonian method, and derivation towards kinematics equations and rigid-body mass matrixes are explained. This nonlinear ROV model represents the underwater thruster dynamics, ROV dynamics and kinematics related to the earth-fixed frame. Multivariable nonlinear adaptive control synthesis using the U-model approach incorporated with radial basis function (RBF) neural networks along with the PID and FLC approaches are implemented using MATLAB� Simulink and integrated with the nonlinear ROV model. Simulations are carried in six degree of freedom (DoF) manoeuvring position in x, y, z coordinates from (0,0,0) to (5,5,1), with the final reference position at (10,10,2). All three controllers are compared and analysed in terms of control synthesis and model tracking capabilities without external disturbances intervention. The simulations are then done with external disturbances intervention for the nonlinear ROV model and the control performances are analysed. The results show good control signal convergence and tracking performance using the U-model control approach. © Science & Technology Research Institute for Defence (STRIDE), 2018.