Hybrid Multi-Gradient Pathfinder (HMGP), a new multi-objective optimization technology is presented which is designed to find the global Pareto frontier and the best Pareto optimal points on this frontier with respect to preferable objectives.HMGP is an evolutionary multi-objective optimization algorithm combined with a gradient-based technique.HMGP is also designed for optimizing very expensive models, and are able to optimize models ranging from a few to thousands of design variables.We investigated HMGP for global multi-objective optimization of reactor core, and compared the results with NSGAII and AMGA.The results indicate that HMGP is more efficient and accurate to achieve the global Pareto frontier solutions.