The New Zealand Cot Death Study, a multicentre case-control study, was set up to identify risk factors associated with sudden infant death syndrome (SIDS). In the 3 years of the study there were 485 infant deaths classified as SIDS in the study areas and 1800 infants who were randomly selected as controls. Data were collected by parent interviews and from obstetric notes. A full set of data for this analysis was available from 356 cases and 1529 control infants. The relationship between length of any breastfeeding and SIDS was examined: 92% of the controls were initially breastfed compared to 86% of the cases. As time went by, cases stopped breastfeeding sooner than controls: by 13 weeks, 67% controls were breastfed versus 49% cases. A reduced risk for SIDS in breastfed infants persisted during the first 6 months after controlling for confounding demographic, maternal and infant factors. Infants exclusively breastfed 'at discharge from the obstetric hospital' (odds ratio [OR] = 0.52, 95% confidence interval (CI): 0.35-0.71) and during the last 2 days (OR = 0.65, 95% CI: 0.46-0.91) had a significantly lower risk of SIDS than infants not breastfed after controlling for potential confounders. We have shown a substantial association of breastfeeding with a lowered risk for SIDS. This supports the need for more positive promotion and active community support to further enhance the level and length of exclusive breastfeeding.