Recent studies of the effects of water and sanitation conditions on infant mortality have produced contradictory findings. This study expands upon previous research to investigate water and sanitation effects in light of other variables: mothers' education, infant feeding practices other than breastfeeding, crowded living conditions, and family income. Seven hypotheses were examined using a linear probability model and data from the 1976-77 Malaysian Family Life Survey. The authors conclude that the health effects of water and sanitation improvements are dependent on child care and feeding practices. Findings indicate that sanitation can greatly reduce infant mortality when mothers are illiterate; that piped water supplies help reduce infant mortality when weaning begins (but not if sweetened condensed milk is the food chosen); and that toilets also have a beneficial effect during the weaning process (most likely as a means of preventing food contamination).