Polygyny is cross-culturally common and a topic of considerable academicandpolicyinterest,oftendeemedaharmfulculturalpractice serving the interests of men contrary to those of women and children. Supporting this view, large-scale studies of national African demographic surveys consistently demonstrate that poor child health outcomes are concentrated in polygynous households. Negative population-level associations between polygyny and well-being have also been reported, consistent with the hypothesis that modern transitions to socially imposed monogamy are driven by cultural group selection. We challenge the consensus view that polygyny is harmful, drawing on multilevel data from 56 ethnically diverse Tanzanian villages. We first demonstrate the vulnerability of aggregated data to confounding between ecological and individual determinants of health; while across villages polygyny is associated with poor child health and low food security, such relationships are absent or reversed within villages, particularly when children and fathers are coresident. We then provide data indicating that the costs of sharing a husband are offset by greater wealth (land and livestock) of polygynous households. These results are consistent with models of polygyny based on female choice. Finally, we show that village-level negative associations between polygyny prevalence, food security, and child health are fully accounted for by underlying differences in ecological vulnerability (rainfall) and socioeconomicmarginalization(accesstoeducation).Wehighlightthe need for improved, culturally sensitive measurement tools and appropriate scales of analysis in studies of polygyny and other purportedlyharmfulpracticesanddiscusstherelevanceofourresultsto theoretical accounts of marriage and contemporary population policy