Comparisons are made between self-reported offenses which may or may not have resulted in being picked up by the police and reports of arrests and convictions. The data presented are based on a sample of 2,312 polydrug-abusing adolescent clients (excluding users of heroin and illegal methadone). Substance offenses were reported by 75.5 percent of the sample, and 44.4 percent reported property offenses, followed by 30.8 percent reporting victim offenses. The study documents that, at least for adolescent drug abusers who enter treatment programs and have not used heroin or illegal methadone, there is a considerable history of criminal offenses involving not only illegal substances, but also crimes against property and person. It has been further shown that variables such as age of first use of an illicit drug, socioeconomic status of the family, number of different drugs used, 'risk level' of drugs used, overdose, and suicide attempts all differentiate offenders from nonoffenders. In a few instances, these variables differentiate the types of offenses committed. It should follow that an increase in the level of drug use, particularly of high-risk drugs, will bring an increase in criminal activity rates. It also follows that success in the prevention or adequate treatment of adolescent drug abuse should have a beneficial effect in reducing the amount of criminal involvement by persons in this age group. Fifteen notes and tabular and graphic data are provided. (Author summary modified)