The overall mortality of acute myocardial infarction (AMI) is still high, half of deaths occurring in the prehospital setting. The American Heart Association (AHA) recommendations strongly emphasize the rapid use of "Emergency Medical Services". Despite this guideline, we frequently observe patients with AMI arriving at the hospital without an emergency ambulance service (SMUR). We undertook a retrospective study to quantify this problem with special interest in the mean of transport, in the role of the primary care practitioner and in the influence of the SMUR on the in-hospital delay. We had 125 AMI in 2000 and 2001: 42 reached the hospital by self-transportation, 57 by ambulance with SMUR and 26 by ambulance without SMUR. An out-hospital doctor was first warned by half of the patients 38% of which arrived by ambulance with SMUR, 37% by ambulance without SMUR and 25% by their own means. A thrombolysis was applied in one third of the AMI: the median prehospital and in-hospital delays were 120 and 40 minutes respectively. The in-hospital delay was significantly shorter when the emergency ambulance service was used. Our results confirm a suboptimal utilisation of the emergency ambulance service for AMI, due to patient and out-hospital doctors. This encourages us to go on providing specific information to the population but also to the primary care practitioners.