Clean rooms are widely used in food-processing industries to guarantee the quality of food products. To maintain the desired cleanliness class in these large and complex enclosures, high ventilation rates are usually used. The recommended air change rates, based on empirical formulas, generate large energy consumptions and do not always guarantee the desired cleanliness class. The objective of this work is to provide a new methodology to properly design the air distribution in clean rooms and to optimise the process arrangement. A complete numerical study is made on the efficiency of transient contaminant removal from a full scale clean room intended for food manufacturing. We combine CFD simulations with the calculation of selected ventilation criteria. The contaminant removal effectiveness and the mean age of air permit to optimise the contaminant source position and to determine decontamination time respectively. These numerical results corroborate previous experimental observations and can lead to new recommendations. (C) 2004 Elsevier Ltd. All rights reserved.