We used PET (15)O and a modified version of the Stroop task to determine if 25-day abstinent heavy marijuana (MJ) users have persistent deficits in executive cognitive functioning (ECF) and brain activity. Performance on a modified version of the Stroop task and brain activity was compared between 25-day abstinent, heavy marijuana users (n = 11), and a matched comparison group (n = 11). The 25-day abstinent marijuana users showed no deficits in performance on the modified version of the Stroop task when compared to the comparison group. Despite the lack of performance differences, the marijuana users showed hypoactivity in the left perigenual anterior cingulate cortex (ACC) and the left lateral prefrontal cortex (LPFC) and hyperactivity in the hippocampus bilaterally, when compared to the comparison group. These results suggest that marijuana users display persistent metabolic alterations in brain regions responsible for ECF. It may be that marijuana users recruit an alternative neural network as a compensatory mechanism during performance on a modified version of the Stroop task. These differences in brain activity may be a common denominator in the evolution of maladaptive behaviors such as substance abuse and other neuropsychiatric disorders.