This paper is concerned with neural-learning control of nonlinear dynamical systems. A variable neural network is introduced for approximating unknown nonlinearities of dynamical systems. Based on variable neural networks, adaptive neural control and predictive neural control schemes are studied. In the adaptive neural control scheme, the weight-learning laws and adaptive controller developed using the Lyapunov synthesis approach guarantee the stability of the overall control system. The convergence of tracking and modelling errors are analysed. The predictive neural control scheme results in simple and easy implementation of nonlinear predictive control. An application of neural-learning control to industrial combustion systems is also discussed.