The results of our effort to avoid the confounding effects of alcoholism on taste preference by studying nonalcoholic children of alcoholic fathers has recently been replicated by other researchers (1). Nonetheless, we agree with Dr. Andrade’s comment that the lack of an association between sweet preference and paternal history of alcoholism does not preclude an association between sweet preference and alcoholism risk. Since studies comparing nonalcoholic groups with and without paternal alcoholism do not directly measure alcoholism risk, the only study design that we believe fully addresses this question is a prospective, longitudinal one. However, such a study would be costly and time-consuming. In view of the recent report of an association between a family history of alcoholism and sweet preference among alcoholics (2), further studies of sweet preference among alcoholics (e.g., after variable periods of abstinence from alcohol) may clarify whether sweet preference is a valid phenotypic marker of alcoholism.