A large number of design guidelines and tools are available for the design of passive cooling systems. Using the underlying thermodynamic models, a certain input (e.g. air change rate, internal heat gains or sun control) results in a certain output (i.e. room temperature). However, in real buildings the room temperature at a given outdoor temperature is a distribution rather than a single value. Therefore, the building engineer should take uncertainties into account, since the actual use of the building, the building physical properties or the user behaviour are statistically distributed. One promising approach to include these uncertainties in the design procedure is the use of statistical models: the design parameter is defined by a mean value and its deviation. From a control theoretical point of view, the deterministic controlled system responds to random disturbance variables by a statistically distributed response function. Considering the institute building of Fraunhofer ISE as example, this study shows how statistical simulations can be applied to the design process of passive cooling in low-energy office buildings. (c) 2006 Elsevier Ltd. All rights reserved.