The three year study of sudden unexpected death in infancy (SUDI), recently completed as part of the programme of the Confidential Enquiry into Stillbirths and Deaths in Infancy (CESDI), includes the largest and most comprehensive study of cot death yet conducted in the UK. Over 450 sudden deaths, of which 80% were classified as sudden infant death syndrome (SIDS) were investigated in five (former) NHS regions between 1993 and 1996, both by confidential inquiry and by case-control technique. Outline results for the first two years were given in the third report of the National Advisory Body for CESDI1 while more detailed reports on various aspects of the case-control study are being published in professional journals.2 3 The national intervention on sleeping position in 1991 was followed by a marked and most welcome fall in the incidence of cot death, the national rate declining from 1.61/1000 live births in 1990 to 0.6 in 1993. However SIDS is still the largest category of deaths in the postperinatal period, and the SUDI study has shown that several major problems remain to be addressed. Some of these I should like to discuss.