The use of anabolic-androgenic steroids (AAS) by young athletes has been a primary concern of sports governing bodies because of the implications for unfair advantage in performance and the potential for adverse side effects. Research over several decades indicated lifetime prevalence of AAS use for adolescent males of 4%-6% and for females of 1.5%-3%, indicating a problem involving millions of athletes and a potential epidemic of AAS-related pathologies. However, recent studies have questioned the presumption that participation in organized sport is the primary risk factor for AAS use in adolescents as well as the extant estimates of the magnitude of the problem. Increasing evidence indicates that AAS use is associated with nonathletes, is linked to a broader syndrome of problem behaviors rather than efforts to achieve sporting success and that sports participation may be protective against AAS use. Moreover, employing lifetime prevalence to gauge AAS use limits accurate evaluation of the personal and public health risk as the majority of respondents are not habitual users. Previous studies may have also inflated prevalence values through ambiguously worded survey questions and other design flaws and few data are available on actual dosages. Prevention efforts need to be focused beyond organized sport and target the general adolescent population rather than athletes and should be founded on interventions with demonstrated efficacy for delinquent, anti-social and self-destructive behaviors rather than the ethical imperative of fair play.