Grid-connected photovoltaic systems have been widely utilized as means of renewable energy-based electricity supply worldwide. Nevertheless, one of the major issues in their implementation is optimal system design, also associated to system sizing. An undersized or oversized system components could compromise the technical benefits of the systems. Therefore, this paper discusses a Grey Wolf Optimization (GWO) for optimizing a grid-connected photovoltaic system design. The optimization problem was devised based on single objective optimization with models of photovoltaic module and inverter set as the decision variables and specific yield transcribed as fitness value that needs to be maximized. Before optimization, an iterative-based sizing algorithm was formulated to determine the optimal module and inverter that give the highest specific yield using a non-computational intelligence approach. Later, GWO was employed to determine the maximum specific yield by choosing the optimal model of PV module and inverter. The results showed that GWO was able to produce same specific yield obtained by iterative-based sizing algorithm. Additionally, GWO was observed to be about 11.3 times faster when compared with the iterative approach. When comparing with particle swarm optimization and genetic algorithm, GWO was discovered to produce higher specific yield with relatively similar computation time.