The Bwamanda hospital insurance scheme in Zaire was launched in the mid-eighties and is one of the few well-established and documented initiatives in the field of district-based insurance schemes in sub-Saharan Africa. It was established that hospital utilization in Bwamanda is significantly higher among the insured population. A higher hospital utilization is however not a goal in itself: it is a positive phenomenon if it takes place for problems where the hospital's know-how and technology are needed to solve the patient's problem. This paper investigates the effect of the insurance scheme on hospital utilization patterns. More specifically, the distribution of this higher utilization over the different hospital departments, as well as its spatial distribution in the entire district area are analyzed. The impact of the insurance scheme on the effectiveness, equity and efficiency of hospital utilization are discussed. The relevance and possible implications of these findings on the design of the Bwamanda insurance scheme are discussed. Finally, it is argued that the methods used in the present study contribute to a coherent framework for the evaluation of similar initiatives.