During a 10-year period, 203 traffic accidents were suspected of being caused by epilepsy; 155 of them definitely caused by seizures were selected for study. The seizures of 75% were psychomotor; in 12% the seizure was the first, and only 4% of the rest had reported that they had epilepsy when they applied for a license. Traffic accidents due to epilepsy were less serious than the "average accident". They involved another vehicle much less often, and they occurred more often outside rather than inside built-up areas. Accidents caused by epilepsy were rare--1/10,000 in the 10-year period. The characteristic features of the traffic accident due to epilepsy may be due to the "random" occurrence of seizures in traffic, and to the seizure-suppressing effect of increased vigilance in city traffic.