Abstract
This paper attempts to defend the theory of risk homeostasis concerning the causation of traffic accidents against critical comments formulated by McKenna (1982). Both methodological and conceptual considerations are presented for this purpose, as well as some additional empirical data which appear to corroborate the theory under debate. Special attention is drawn to the importance of the choice of denominator for the calculation of accident rates (per km, per h or per capita) in studies aimed at an evaluation of the effectiveness of accident countermeasures.