Photovoltaic (PV) energy generation is one of the more promising renewable energy technologies. However, determining optimal specifications (e.g., size, location) for PV systems is a significant challenge for planners and investigators of distribution systems. In this paper, a framework is provided for integrating geographical information systems, artificial bee swarm optimization, and simulation to determine the optimal size and location for PV panels on an annual basis. A geographical information system module is developed first to determine the suitable location and capacity for a stand-alone PV system in rural areas, taking into account the technical, environmental and socio-economic criteria affecting the site selection process of stand-alone PV systems. Then an efficient heuristic approach, namely artificial bee swarm optimization, is proposed for optimum sizing of a stand-alone PV system to continuously satisfy the load demand with minimal cost and enhanced energy system reliability. For this aim, the maximum allowable loss of power supply probability is also considered. The proposed framework is applied to a real case study in eastern Iran and the results show that it can effectively be used for off-grid photovoltaic projects in rural areas considering both power quality and cost.