This paper introduces a Binary Particle Swarm Optimization based method to accomplish optimal location and size of a Photovoltaics Grid-Connected System (PVGCS) for distributed power generation. The main technical constraint is the maximum installed peak power, which is limited for utilities Power Distributor Company. The fitness function to be optimized is the profitability index. A fair comparison between the proposed algorithm and other methods is performed. For such goal, convergence curves of the average profitability index versus number of iterations are computed. The proposed algorithm reaches a better solution than genetic algorithms when considering similar computational cost (similar number of evaluations). (C) 2010 Elsevier Ltd. All rights reserved.