The leading substance of misuse among teens after tobacco, alcohol, and marijuana is the use of pain killers. Very few longitudinal studies on prediction of pain killer use have been conducted among teens. This study examined the 1-year prediction of self-reported last 30-day pain killer use controlling for baseline 30-day painkiller use among 1186 alternative high school youth in California. Among demographic, behavioral, psychosocial, and environmental predictors, a multivariable model indicated that: (a) relatively higher levels of baseline pain killer use; (b) white ethnicity; (c) relatively lower levels of depressive symptoms (contrary to previous studies); and (d) those who live with both parents were more likely to report use of pain killer medications in the next year. It is speculated that those with relatively greater access to pain medication, within an at-risk social environment, are those who will use it later on. (Contains 3 tables.)