The relationship between the duration of breast feeding and psychosocial outcomes measured between the ages of 15 and 18 years was examined in a birth cohort of 999 New Zealand children. During the period from birth to 1 year, information was collected on maternal breast-feeding practices. Between the ages of 15 and 18 years, sample members were assessed using a range of psychosocial measures, including measures of the quality of parent-child relationships, juvenile delinquency, substance abuse and mental health. Children who were breast fed for a longer duration were more likely to report higher levels of parental attachment and tended to perceive their mothers as being more caring and less overprotective towards them compared with bottle-fed children. No association was found between the extent of breast feeding and subsequent rates of juvenile offending, substance use and mental health in later life. Mothers who elected to breast feed were also more likely to be older, better educated, living with a partner, less likely to smoke during pregnancy, and to come from advantaged socio-economic backgrounds characterised by better living standards and a higher family income. Rates of breast feeding were also greater among mothers who gave birth to a first-born child of higher birthweight. After adjustment for these maternal and perinatal factors, the duration of breast feeding remained significantly associated with adolescent perceptions of maternal care, with increasing duration of breast feeding being associated with higher levels of perceived maternal care during childhood. It is concluded that: (a) it is unlikely that breast feeding is associated with reduced risks of psychiatric disorder in later life; (b) breast feeding may lead to closer parent-child relationships; and (c) it is unlikely that the association between breast feeding and cognitive development is mediated by intervening processes relating to improved psychosocial adjustment in breast-fed children.