Parents in many traditional and industrialized societies have been seen to invest differentially in given offspring according to the total pool of resources available, as well as the offspring's age, gender, and perceived vigor or condition. The discipline of evolutionary ecology offers a framework for understanding these differential parental investment strategies. Evolutionary models of parental investment predict that parents adjust the magnitude and scheduling of resources allocated to their children according to the likely fitness payoffs of current investments and those from alternative uses of resources. To date, however, little research has attempted to test hypotheses explicitly derived from this body of theory. This project seeks to test several predictions of evolutionary parental investment theory using data from an observational study of mother-offspring proximity, and breast-feeding and supplementation practices among the Au of Papua New Guinea. Prolonged lactation leads to a protracted anovulatory period as well as to long-term declines in maternal energy reserves which have been documented to affect maternal health and survival. As such, lactation is one of the costliest forms of investment expended by mothers on their newborn offspring and should vary systematically with individual characteristics of the mother and child. This study will yield important data on variation in infant feeding patterns in a traditional society and will help to identify the conditions under which specific children are at elevated risk of growth faltering, morbidity, and death. It will also provide one of the first critical tests of evolutionary parental investment theory.