AbstractOn the basis of questionnaires administered to almost 2, 000 high school students in Cali, Colombia, a subset of items was selected that deal primarily with parent-child relationships. This 53-item set, referred to as the Drug Risk Scale (DRS), was administered to two new cross-validation samples, one consisting of high school students and the other consisting of drug addicts attending drug rehabilitation centers. Significant differences in parent-child relations were found between these new groups. The DRS was also found to have reasonably high sensitivity and specificity. Its potential value as a risk-prediction instrument is discussed