The obstructive sleep apnoea syndrome (OSAS) is a condition causing daytime sleepiness and has been related to an increased risk for traffic accidents. However, the evidence linking severity of OSAS to a higher rate of automobile crashes is based on limited data. The aims of this study were to study the traffic accident rate in the last 5 years in patients referred to our sleep clinic because of clinical suspicion of OSAS and to analyse variables related to an increased risk for traffic accidents. A series of 189 consecutive patients with a driving license referred for a sleep study because of OSAS clinical suspicion and a control group (CG) of 40 hospital staff workers who denied snoring, matched for age and sex with the study population, were studied. Patients underwent a full-night polysomnography and both patients and the CG completed a self-answered questionnaire. One hundred and twenty-two patients were diagnosed as OSAS and 67 patients as non-apnoeic snorers (NAS). The self-reported number of accidents was significantly higher in OSAS patients compared with CG. The self-reported number of times off the road was significantly higher in OSAS patients compared with NAS and with CG. Variables associated with an increased risk for traffic accidents were self-reported sleepiness while driving (OR 5, 95%CI 2.3-10.9), having quit driving because of sleepiness (OR 3, 95%CI 1.1-8.6) and being currently working (OR 2.8, 95%CI 1.1-7.7). We conclude that self-reported sleepiness while driving is associated with an increased risk for traffic accidents in OSAS patients and in NAS. We suggest that this symptom can be used to alert patients and to give priority in the sleep clinic for study and treatment.