The emergency medical service in France (called SAMU-Service d'Aide Médicale Urgente) is a public service responsible for provision of basic life support to all emergency requests, essentially external from health units. This paper applies discrete simulation techniques to study the configuration and operation of a SAMU in France, analyzing the performance of the system and investigating alternative configurations and dispatching policies for its operations. In particular, we evaluate the fraction of calls are responded within a threshold time T (varying from 10 to 25 minutes), with special concern to cardiac arrest emergencies, and the utilization rates of medical equips. It is show that, modifications on the configuration such as decentralization of medical equips, and modifications on the dispatching can result in significant improvements for the performance measures under analysis. 1 Introduction The emergency medical services (EMS) are considered efficient if they arrive in the emergency requests location as rapid as possible transporting specialized personal and equipment (e.g.; doctor, rescuers, medicines, oxygen…). In additional, an ambulance must be able to transport the patient to a suitable hospital or clinic. Conversely, EMS managers must often to balance the benefits of providing a suitable care as quickly as possible and their costs related to medical resources and capacity (more ambulances and specialists, stations, better training of personnel, improved equipments) in the system. In France, the main pre-hospital care operations external of health care units (hospitals, clinics) are under the management of the SAMU. One of the major concerns of these systems is related to response time (interval between the arrival of the request and the arrival of medical equip at the site where the emergency occurs). As mentioned in [8], other performance measures to an EMS are: the balance of ambulance workloads, the fraction of calls not serviced by the system (loss probability), and the fraction of calls serviced within a predetermined threshold (i.e., fraction of calls with response times not exceeding T minutes). In particular, one of the critical urgencies for the SAMU is the cardiac arrest. In this case the response time, and consequently the reasonable delay to restore the blood circulation is in order of few minutes, otherwise it can result in permanent cerebral sequels or death to the patient. One * This paper was not presented at any other revue. Corresponding author A.P.; Iannoni. Tel. +33 166997762.