In this paper, a neural network IC-based variable step size MPPT controller for fuel cell power system is proposed. The efficiency of the proposed neural network MPPT has been successfully studied using a 7 kW PEMFC supplying a resistive load via a DC DC boost converter derived using the proposed MPPT controller. Obtained simulation results from using Matlab/Simulink environment show that the proposed neural network IC-based variable step size MPPT outperforms the conventional fixed and variable step size MPPT controllers in term of convergence speed, overshoot and steady state oscillations. In addition, the current ripple reduction using the proposed neural network MPPT controller allows to improve the efficiency of the fuel cell and extend its lifetime. (C) 2016 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.