Epidemiological research on psychoactive substance use disorders can and sometimes does lead to conclusions that are mutually conflicting. A current example is the apparent drop in substance use among adolescents in schools, with continuing high levels of substance-associated adolescent suicide, hospitalization, and crime. In an effort to clarify such differences, the author underscores the importance of understanding the three most common epidemiological methods. Each has different histories, advantages, liabilities, inherent value sets, and disparate-but-overlapping purposes.