This report is a product of research conducted under a program exploring improved methods of acquiring data on sensitive topics, such as the scope and intensity of current social problems. In this case, the problem studied was drug abuse in the Army. A mail survey dealing with illicit drug use was conducted, comparing the data acquisition effectiveness of the Randomized Inquiry (RI) technique and a conventional-type questionnaire. The respondents were four stratified random samples, of approximately 500 men each, drawn from the U.S. Army, Continental United States. Variables studied were (a) method used in questionnaire (RI vs. conventional), (b) rank of respondent (officers vs. enlisted men), and (c) effect of advance notice on return rate. Five sensitive questions were developed to get information of the respondents' use of five drugs. Conventional questionnaires yielded more response than questionnaires employing the RI technique; drug usage rates reported were not significantly different under the two techniques. The questionnaire return rate proved to be higher for officers than for enlisted men, particularly when they received advance notice of the survey. Since the research performed for this project is primarily methodological, this report is of interest especially to personnel in the fields of operations research, data acquisition, management, and program planning. (Author)