131 cases of the Sudden Infant Death Syndrome (SIDS) in the municipality of Copenhagen 1956--71 (incidence 0.92 SIDS cases per 1 000 live births) were investigated on the basis of police reports and infant health visitor's records. Fewer SIDS cases were breast-fed than controls from the second week of life to four months of age. No significant differences were found with respect to the concentrations of fresh cow's milk dilutions, age at introduction of solid food, or number of meals per day. From 1956 to 1971 breast-feeding was declining and solids introduced earlier, while the incidence of SIDS remained constant. The lifespan for SIDS cases who had never been breast-fed was equal to that of cases who had. There is no evidence that SIDS victims had their first solid food during the last days of life. The results are discussed in the light of the hypersensitivity, immuno-incompetence, and high solute feeding hypotheses. It is concluded that feeding does not seem to be responsible for the occurrence of SIDS. The lower frequency of breast-feeding among SIDS cases is ascribed to various factors known to be associated with SIDS.