This case report augments a paper published in 1971 (Gunn and Fenton) in which it was indicated that automatic behaviour is a rare explanation for the crimes of epileptic patients. It was claimed that although two possible ‘automatic’ crimes were committed by two epileptic patients among the 46 male epileptics at Broadmoor there were no such crimes committed by any of the 158 male epileptic prisoners who came into a national sample. Since then it has become clear that one man serving life imprisonment, excluded from the epileptic prisoner sample in 1967 because of a doubt about his diagnosis, is definitely epileptic and probably killed his wife during an epileptic attack or its immediate sequela.