BACKGROUND: Cocaine use disorder is associated with cognitive deficits that reflect dysfunctional processing across neural systems. Because there are currently no approved medications, treatment centers provide behavioral interventions that have only short-term efficacy. This suggests that behavioral interventions are not sufficient by themselves to lead to the maintenance of abstinence in patients with cocaine use disorder. Self-control, which includes the regulation of attention, is critical for dealing with many daily challenges that would benefit from medication interventions that can ameliorate cognitive neural disturbances. METHODS: To address this important clinical gap, we conducted a randomized, double-blind, placebo-controlled, crossover design study in patients with cocaine use disorder (n = 23) and healthy control participants (n = 28). We assessed the modulatory effects of acute atomoxetine (40 mg) on attention and conflict monitoring and their associated neural activation and connectivity correlates during performance on the Eriksen flanker task. The Eriksen flanker task examines basic attentional processing using congruent stimuli and the effects of conflict monitoring and response inhibition using incongruent stimuli, the latter of which necessitates the executive control of attention. RESULTS: We found that atomoxetine improved task accuracy only in the cocaine group but modulated connectivity within distinct brain networks in both groups during congruent trials. During incongruent trials, the cocaine group showed increased task-related activation in the right inferior frontal and anterior cingulate gyri, as well as greater network connectivity than the control group across treatments. CONCLUSIONS: The findings of the current study support a modulatory effect of acute atomoxetine on attention and associated connectivity in cocaine use disorder.