Abstract
This article describes a study which was conducted in Detroit, Michigan, to determine the number and location of ambulances necessary to provide adequate service in two selected representative urban areas. A mathematical model was developed to quantify the qualitative effects of alternative recovery system configurations, thereby using data on the characteristics of emergency occurrence and service processes for the areas under consideration. The data analysis and the development of the model are summarised after which the predictions derived from the model are discussed, and the resulting recommendations are presented.