We report a study of sudden unexpected infant death (SUID) in a French county (Seine-Maritime) between 1978 and 1981 (rate of 2.71 per 1000 live births). The results agree with those of previous studies concerning sex ratio, seasonal variation, and peak age of death. No relation between gestational age and age of death was found. A comparison with matched living control infants was performed for 136 of the 207 total deaths for whom computerised medical data were available. Significant differences were found between these two groups in gestational age, birthweight, admission to a paediatric ward at birth, and proportion of single mothers. In analysing birthweight and gestational age separately, the birthweights of SUID cases were significantly lower than those of controls, whatever the gestational age. The converse was not true, however: in classifying infants by birthweight, a significant difference between the gestation period of SUID victims and controls was found only for low birthweight infants (less than 2500 g). It is considered that it may be possible to identify infants at particular risk of SUID within the group of babies admitted to hospital at birth for major problems, and a study is currently underway among infants born in Seine-Maritime in 1982.