In a study of polygyny and female fitness among the Dogon of Mali, West Africa, the author attempts to overcome the methodological problems of studies by measuring female fecundability and child mortality prospectively. She measures female fecundability from two years of data on menstrual hut attendance, as corroborated by hormonal profiles. The author also measures child mortality from eight years of data on child survivorship, analyzed by logistic regression. The results indicated that the odds of death were 7 to 11 times higher under polygyny even in the presence of controls for confounding risk factors. The author presents two additional logistic regression models that implicate polygyny as a predictor of mortality. Polygyny is more likely to adversely affect female fitness when females have low status. To assess the effect of polygyny on aspects of paternal behavior that were both quantifiable and observable, the author compares observations of direct childcare performed by polygynously and monogamously married men.