The uncertainty in long-term photovoltaic (PV) system yield predictions was examined by statistical modeling of a hypothetical 10 MW AC, c-Si photovoltaic system in Toronto, Canada. The goal of this research was to develop a methodology for estimating the uncertainty in PV yield predictions and to identify avenues for reducing this uncertainty. In this case study, uncertainties were estimated to be about 3.9% for year-to-year climate variability, 5% for long-term average horizontal insolation, 3% for estimation of radiation in the plane of the array, 3% for power rating of modules, 2% for losses due to dirt and soiling, 1.5% for losses due to snow and 5% for other sources of error. Uncertainties due to ageing and system availability were also considered. By performing statistical simulations with the Solar Advisor Model software, it was found that the combined uncertainty (standard deviation) is approximately 8.7% for the first year of operation, and 7.9% for the average yield over the PV system lifetime. While these numbers could vary significantly from one system to the next, the methodology developed is widely applicable. Moreover, a simpler methodology was also explored which should yield quick and fairly reliable estimates of uncertainty. Finally, avenues for reducing yield uncertainties were identified, including: increasing the reliability and resolution of solar radiation estimates, including measurements of irradiance in non-horizontal planes at high quality ground stations, reducing the uncertainty in module ratings and investigating losses that have not been well documented such as those due to dirt, soiling and snow. Crown (C) 2013 and Elsevier Ltd. All rights reserved.