A modeling-control paradigm to regulate output of the hippocampus (CA1) for a hippocampal neuroprosthesis was developed and validated using an in vitro slice preparation. Our previous study has shown that the VLSI implementation of a CA3 nonlinear dynamic model can functionally replace the CA3 subregion of the hippocampal slice. The propagation of temporal patterns of activity from DG-->VLSI-->CA1 reproduces the activity observed experimentally in the biological DG-->CA3-->CA1 circuit. In this project, we incorporate an open-loop controller to optimize the output (CA1) response. Specifically, we seek to optimize the stimulation signal to CA1 using a predictive dentate gyrus (DG)-CA1 nonlinear model (i.e., DG-CA1 trajectory model) and a CA1 input-output model (i.e., CA1 plant model), such that the ultimate CA1 response (i.e., desired output) can be first predicted by the DG-CA1 trajectory model and then transformed to the desired stimulation intensity through the CA1 inverse plant model. Laguerre-Volterra kernel model for random - interval, graded - input, contemporaneous - graded -output system is formulated and applied to build the DG-CA1 trajectory model and the CA1 plant model. The inverse model to transform desired output to input is also derived and validated. We validated the paradigm in hippocampal slices, and results showed the CA1 response evoked by the controlled stimulation signal reinstated the CA1 response evoked by the trisynaptic pathway.