Abstract This study was carried out in response to reports from nurses to a post-neonatal mortality review committee that a number of mothers of infants dying from sudden infant death syndrome (SIDS) appeared to be depressed before the child's death. The New Zealand Cot Death Study was a 3 year multicentre case-control study for SIDS. There were 485 SIDS cases in the post-neonatal age group in the study regions, and these were compared with 1800 control infants. Infants of mothers with either a self-reported use of medication for psychiatric disorders, a history of hospitalization for psychiatric illness or a family history of postnatal depression had a significantly increased risk of SIDS compared with infants of mothers who were either not using medication (odds ratio (OR) = 1.45; 95% confidence interval (Cl) = 1.03, 2.04) or were without a history of hospitalization for psychiatric illness (OR = 1.80; 95% Cl = 1.03, 3.11) or a family history of postnatal depression (OR = 1.61; 95% Cl = 1.06, 2.43). All mothers of infants born in the study areas over a 1 year period were eligible to complete a questionnaire measuring maternal depression when the infant was 4 weeks of age. Thirty-three infants subsequently died from SIDS, and they were compared with 174 controls. Fifteen (45.5%) of the mothers of cases were depressed, compared with 28 (16.1%) of the mothers of controls. This prospective study found that the infants of those mothers that were depressed were more likely to die from SIDS than those of the non-depressed mothers (OR = 4.35; 95% Cl = 1.82, 10.37) and postnatal depression as a risk factor for SIDS was still significant after controlling for possible confounding variables (OR = 3.37; 95% Cl = 1.24, 9.12). We conclude that postnatal depression is a risk factor for SIDS.