The hypothesis of a negative association between rates of suicide and number of children in marriage was investigated in a prospective study of 989,949 women followed up for 15 years (1970 through 1985) with 1190 deaths from suicide. Women who had never married exhibited higher relative risks for suicide than married parous and married nonparous women for all age groups younger than 65 years at the start of follow-up. Among the married, the parous women had lower relative risks than nonparous women for all ages. For both premenopausal and postmenopausal women, a strong linear decrease in relative risk for suicide with increasing number of children in marriage was found. The effect of number of children was independent of social class measured as years of completed schooling. The findings provide the first empirical support for theories of parenthood and suicide advanced by Durkheim almost 100 years ago.