The NSDUH is administered to approximately 70,000 individuals annually and oversamples adolescents so that approximately one-third (n=23,300) of the sample includes 12 to 17 year olds. Publicly available state-level estimates for NSDUH respondents aged 12–17 were compiled from 2002–2008 with 2008 being the most current results presently available [9]. The NSDUH survey was initiated in 1999, but due to improvements in data collection procedures in 2002, prevalence rates before 2002 are not comparable and are thus not included [10]. State-level estimates are derived using a survey-weighted hierarchical Bayes estimation approach [10] aggregated over two-year intervals. We analyzed two NSDUH outcomes: (1) prevalence (percent) of past-month marijuana use and (2) percent responding that “great risk” is associated with smoking marijuana at least once a month.
