A longitudinal study of 1,089 epileptic patients followed up by the same specialist between 1965-1991, allowed close observation of the seizures occurring to the patient at the wheel and their consequences and to relate them to detailed epileptological criteria. The results show road accidents caused by epileptic seizures are few and most of them are minor. The repatriation of risks between patients is very uneven. The quality of the neuro-psychic inter-critical state as well as the patients' degree of compliance seem to be more reliable risk indicators than some more traditional criteria like the length of remission between seizures. Although seizures occur more frequently in patients suffering from Complex Partial seizures as opposed to other forms of epileptic seizures, the differences between patients with epilepsy lies mostly in their behaviour and in their own representation of the risks. There is a need for a body of rules and regulations serving as an official framework regulating the driving test. This widely circulated document should take into account the multiplicity of cases, including the small number of patients thought to be dangerous. Its mode of application should allow doctors as well as patients to opt for a realistic attitude based on decision-making criteria involving a thorough knowledge of epilepsy as well as a thorough knowledge of the psychological characteristics of the patient concerned.