The treatment of brain arteriovenous malformations (AVMs) has been studied extensively. With the use of the Spetzler-Martin (S-M) grading system, patients can be informed appropriately about their possible surgical risks. This does not hold true for their neuropsychological sequelae, which have not been studied widely. We evaluated the neuropsychological outcome of our patients treated for brain AVMs. Of 113 patients treated for a brain AVM between 2001 and 2009, 66 patients were enrolled in the study. All patients underwent treatment at our institution and neuropsychological testing 2 years later using a test battery constructed specifically for this study. A control group consisted of 10 subjects without any neurologic disease. When the whole cohort was analyzed, no significant differences were found between the groups distinguished by hemorrhage, gender, or hemispheric dominance. Patients with S-M IV and V scores fared significantly worse than patients with S-M I to III. Patients who presented with epilepsy scored lower than patients presenting with other symptomatology, but the difference had only borderline significance. When we analyzed patients according to the presence or absence of obliteration after treatment and compared these with the control group, we found no significant differences. When the patients with an obliterated AVM after treatment were subdivided according to treatment modality, there were no significant differences in their S-M groups compared with the control group. Similarly, those patients with nonobliterated AVMs analyzed according to their S-M grade showed a borderline significant difference, with S-M IV and V having a worse neuropsychological outcome compared with the other groups. Patients harboring nonobliterated high-grade AVMs (S-M IV and V) scored worse than patients with nonobliterated AVM S-M grades I to III. This could be explained by the steal phenomenon. No differences in neuropsychological testing were found when comparing results according to nidus location. This study lends support to an active treatment policy for cerebral AVMs. Those patients in whom complete obliteration was achieved with treatment scored similarly to the background population, implying active AVM treatment does not cause deterioration in neuropsychological performance. This, together with a> 90% AVM obliteration rate, favors microsurgery as the treatment modality of choice whenever the AVM can be safely resected.