The performance of a model-based control system depends strongly on the accuracy of the process model used. Many real-time processes have uncertain and non-linear dynamics and so it is difficult to model them mathematically. Because of the function approximation ability of neural networks, much research has been conducted on adapting them for modeling and controlling dynamic systems. The main objective of this paper is to develop fuzzy neural schemes in the field of process identification and control. The paper develops a systematic and transparent non-linear Fuzzy-Neural Model Predictive Control (FNMPC) scheme to inherit the advantages of both fuzzy logic systems and neural networks and to avoid their individual shortcomings. A reason for this scheme is that most of the optimization methods previously used in Model Predictive Control (MPC) were very computationally demanding. Consequently, the costing horizon of such MPC schemes is limited to a few time-steps and hence the actual control actions are erratic. For comparison reasons, both Fuzzy MPC (FMPC) and FNMPC have been simulated on the mean arterial blood pressure control problem.