A systematic comparison of the predictive accuracy of simplified design methods with respect to the relief diameter and the discharged mass flow on the basis of experimental results obtained in reactors with volumina ranging between 0.01 and 10 m(3) allow for the conclusion that empirical design rules are not suited for an extrapolation. The preliminary recommended method in case of vapour systems is this of J.C. Leung and for use with gassy or hybrid systems the method by H.K. Fauske and J.C. Leung is promoted. Indeed, partly non acceptable overpredictions are evident. By using SAFIRE/VENT reasonable predictions are possible though still an oversizing can not be cut off. (C) 2000 Elsevier Science Ltd. All rights reserved.