Hormonal factors have been suggested as a cause of post-natal depression, but suckling frequency which is the major influence on hormone levels post-partum has not been considered in previous studies. Eighty-nine women who had taken part in a prospective study of post-natal depression were asked about their feeding patterns 18 months after childbirth. Half the mothers who were breast feeding introduced solids or artificial milk feeds before 12 weeks and were likely to have lowered prolactin levels and increased ovarian follicular activity. Mothers who totally breastfed their babies for at least 12 weeks or who were on the pill, had a higher incidence of post-natal depression than those who were not on the pill or who partially breastfed. Among the 62 women who attempted breast feeding those most likely to have normal levels of endogenous hormones were those least likely to have depressive symptoms.