The French Emergency Medical service, known as SAMU, is responsible for providing permanent phone support and dispatching the proper response for emergency requests. The response time required for an ambulance's arrival at the scene following a call is an important performance indicator in determining the quality of the SAMU system since this may be directly related to patient's survival. In this paper, discrete simulation techniques are used to model the SAMU of the Val-de-Marne department (France) in order to investigate several alternative configurations for potential improvements. Scenarios consist of adding more resources, relocating existing teams and reducing processing times in order to improve response time. We found that repositioning part of the existing teams into potential stations increased average percentage of calls covered within the 20-minutes criterion up to 4.8%. This improvement in coverage reaches 5.2% when reducing the regulation processing time by 20%.