Converging evidence suggests that hyperactivity in frontal-striatal circuits and in action-monitoring processes characterizes obsessive-compulsive disorder (OCD). It remains unclear, however, just how these abnormalities in brain function translate into the cognitive, affective, and behavioral manifestations of OCD. One possibility is that exaggerated or false error signals generated by the anterior cingulate (ACC) underlie compulsive behaviors by triggering the feeling that things are "not just right" even when no actual error has been made. Since recurrent compulsive behavior typically follows correct completion of a behavioral task (e.g., hand washing), ACC hyperactivity should be observed during correctly completed, high-conflict trials as well as during error trials. Frontal-striatal regions would also be expected to be activated during both trial types, as these regions are robustly associated with OCD across multiple neuroimaging paradigms. To test this hypothesis, 14 OCD patients and 14 matched controls completed a speeded reaction time task during functional magnetic resonance imaging (fMRI). Only correctly rejected, high-conflict trials produced excessive activation in both action monitoring (rostral and caudal ACC, LPFC) and frontal striatal regions (lateral orbitofrontal cortex (OFC), caudate, and thalamus) among OCD patients when compared to healthy controls. Portions of the posterior cingulate were also hyperactive among OCD patients. These results suggest that correctly rejected, high-conflict trials that require response inhibition may provide a better model than error trials of compulsive behaviors in OCD.