The study used a mixed cross-sectional and longitudinal design to test whether a behavioral choice task, previously validated in stimulant users, was associated with increased opioid misuse severity at baseline, and whether it predicted change in opioid misuse severity at follow-up.