This paper investigates the effects of weather forecast uncertainty on the performance of a concept for predictive control of building systems operation. The concept uses a computational physically-based building model and weather forecasts to predict future heating or cooling requirement. This information enables the building systems to respond proactively to keep the operational temperature within the thermal comfort range with the minimum use of energy. The effect of weather forecast uncertainty was assessed using weather data from two different years in a temperate climate in the simulation of 24 building design scenarios. Despite the uncertainty in the weather forecasts, the predictive control concept demonstrated a potential for energy savings and/or improvements in thermal indoor environment when compared to a conventional rule-based control. (C) 2013 Elsevier Ltd. All rights reserved.