This paper proposes a Hybrid Nelder-Mead and Cuckoo Search (HNMCS) algorithm to minimize the power loss in hybrid AC/DC microgrid systems by optimizing the output power of Renewable Energy Distributed Generators (REDG). The non-linear power loss minimization problem is solved by the proposed HNMCS to optimize the size of REDG. So far, the REDG sizing is determined by considering generator output as variable whereas in the proposed technique, the area required for the operation of REDG in hybrid AC/DC microgrid is taken as variable. The microgrids are developed by categorizing the existing distribution system to multiple zones. A hybrid AC/DC microgrid is developed with AC grids supported by substation and DC grids operated by their individual REDG units. The suitable location for REDG units including the combination of solar-photovoltaic modules and fuel cells in DC grid is identified by Loss Reduction Sensitivity Factor (LRSF). A standard 33-bus and 69-bus radial distribution system is modeled as a hybrid AC/DC microgrid system. The system is analyzed for its performance in stand-alone system and extended to zone cataloging as residential, industrial and commercial zones. The proposed HNMCS algorithm identifies the optimal solution for REDG sizing with improved convergence rate and reduced simulation time. (C) 2018 Elsevier Ltd. All rights reserved.