Persistent dose-related cognitive decrements have been reported in 28-day abstinent heavy marijuana (MJ) users. However, the neural substrates of these decrements in cognitive performance are not known. This study aimed to determine if 25-day abstinent MJ users show persistent dose-related alterations in performance and brain activity using PET H(2)(15)O during the Iowa Gambling Task-IGT (a decision-making task). Eleven heavy MJ users and 11 non-drug users participated. The MJ group resided in an inpatient research unit at the NIH/NIDA-IRP for 25 days prior to testing to ensure abstinence. A dose-related association was found between increased MJ use and lower IGT performance and alterations in brain activity. The MJ group showed greater activation in the left cerebellum and less activation in the right lateral orbitofrontal cortex (OFC) and the right dorsolateral prefrontal cortex (DLPFC) than the Control group. When the MJ group was divided into Moderate (8-35 joints/week) and Heavy users (53-84 joints/week), the Heavy MJ group showed less activation in the left medial OFC and greater activation in the left cerebellum than the Moderate group. However, brain activity and task performance were similar between the Moderate MJ users and the Control group, suggesting a "threshold effect". These preliminary findings indicate that very heavy users of MJ have persistent decision-making deficits and alterations in brain activity. Specifically, the Heavy MJ users may focus on only the immediate reinforcing aspects of a situation (i.e., getting high) while ignoring the negative consequences. Thus, faulty decision-making could make an individual more prone to addictive behavior and more resistant to treatment. Finally, it is unclear if these neurologic findings will become progressively worse with continued heavy MJ use or if they will resolve with abstinence from MJ use.