Recent epidemiologic evidence suggests that children in developing countries who terminate breast-feeding before their first birthday are less likely to be malnourished than those who are breast-fed for longer durations. This finding calls into question the advice women are given to breast-feed as long as possible. This review examines several studies that found a relationship between prolonged breast-feeding and malnutrition. Many studies have shown a negative association between prolonged breast-feeding and growth, but there is little reason to expect the association to be causal. Problems in study design and analysis, such as failure to control for confounding and reverse causality, leave the nature of the association largely unspecified.