The Randomized Response Technique (RRT) appears to have promise in future work which studies the relation of school variables to disruption or delinquent behavior. The RRT is especially useful in situations when it is difficult or undesirable directly to ask stigmatizing questions. The proportions of students in this study estimated to have used morphine, marijuana or cigarettes were generally higher when self-report techniques were used. Use of RRT achieves confidentiality by introducing random error into individuals' responses, and consequently offers more potential for estimating proportions, testing for differences between proportions, and doing time series studies of delinquency. (BN)