OBJECTIVES--To calculate age adjusted mortality ratios for suicide by women in the first year after childbirth and during pregnancy, and to identify characteristics of postnatal suicide in childbearing women. DESIGN--Retrospective study based on population data for England and Wales from 1973 to 1984. SUBJECTS--Women aged 15-44 who committed suicide in the year after childbirth or during pregnancy. MAIN OUTCOME MEASURES--Observed to expected mortality ratios for the total postnatal sample, for five year age groups, for unmarried mothers, and for suicide after stillbirth; observed to expected mortality ratios for all suicides during pregnancy and for five year age groups; the timing of suicides in relation to delivery; and the frequency of the various methods of suicide. RESULTS--The standardised mortality ratio for postnatal suicide was calculated to be 0.17--that is, the actual total was only one sixth of that expected. The low ratio was not found after stillbirth, which was associated with a rate six times that in all women after childbirth. The low ratio was less pronounced, but still present, in teenage mothers and in unmarried mothers. Women who committed suicide after childbirth most often did so in the first month, and there was a tendency to use violent methods. The standardised mortality ratio for suicide during pregnancy was calculated to be 0.05 of all pregnant women. Teenagers were at substantially increased risk. CONCLUSIONS--Women in the first year after childbirth and during pregnancy have a low risk of suicide despite their high rate of psychiatric morbidity. Underreporting of maternal suicides is unlikely to explain the findings, though it may affect their magnitude. Motherhood seems to protect against suicide. Concern for dependants may be an important focus for suicide prevention in clinical practice.