In this paper, an intelligent control scheme using a fuzzy single neuron controller (FSNC) is proposed for nonlinear process control. Different from the conventional three-parameter single neuron controller (SNC), the FSNC uses the fuzzy logic concept to build its nonlinearity rather than just mapping through a nonlinear saturation function. This effort, though it introduces additional tuning parameters, can greatly enhance the nonlinear capability for process control. A simple yet efficient parameter tuning algorithm has been developed, which enables the FSNC to learn to control the nonlinear process adaptively with merely observing the process output errors. Both the convergence of the proposed parameter tuning algorithm and the stability of the presented FSNC-based control system are guaranteed by utilizing the Lyapunov stability theorem. To demonstrate the applicability and effectiveness of the proposed intelligent control scheme, we apply it to the direct adaptive control of an unstable nonlinear CSTR. Comparisons with an adaptive SNC and PID controllers were performed. Extensive simulation results reveal that the proposed FSNC is promising and is well suited to the direct adaptive control of processes having severe nonlinearity.