This paper addresses the tracking control problem of roboticmanipulators with unknown and changing dynamics. In this study, nonlineardynamics of the robotic manipulator is assumed to be unknown and a controlscheme is developed to adaptively estimate the unknown manipulator dynamicsutilizing generic artificial neural network models to approximate the underlyingdynamics. Based on the error dynamics of the controller, a parameter updateequation is derived for the adaptive ANN models and local stability propertiesof the controller are discussed. The proposed scheme is simulated andsuccessfully tested for trajectory following tasks. The controller alsodemonstrates remarkable performance in adaptation to changes in manipulatordynamics.