In this aging society, attention has not been fully given to brain arteriovenous malformations (AVMs) in elderly patients. This study sought to describe a single institution's experience treating arteriovenous malformations (AVMs) in elderly patients. We conducted a retrospective review of brain AVMs in elderly patients treated at our institution between 1990 and 2012 with a focus on the clinical features, risk of hemorrhage and treatment outcomes. Of the 2790 patients in our AVM database, 98 patients were over the age of 60 at presentation. Forty-eight percent presented with hemorrhage. Risks of initial hemorrhage were history of hypertension, smaller AVM size (< 3 cm) and exclusively deep venous drainage. Treatment modalities were microsurgical resection in 65 %, embolization alone in 10 %, stereotactic radiosurgery (SRS) in 11 % and observation in 14 %. Preoperative embolization was performed in 32 % in the surgical group. Complete obliteration was achieved in 95 % by microsurgery, 30 % by embolization alone and 45 % by SRS. Good functional outcome (modified Rankin Scale, mRS < 2) was achieved in 69 % after a median follow-up of 5.8 years. Multivariate logistic analysis revealed that a pretreatment mRS score a parts per thousand yen2, eloquent location and higher S-M grade (IV or V) were associated with worsening functional status, whereas surgical resection was a negative factor. Posttreatment hemorrhage occurred in 8 %. AVM-related death occurred in three patients (2 by surgery and 1 by observation). Brain AVMs in elderly patients still pose a high risk of hemorrhage. Initial hemorrhage may be associated with a history of hypertension, AVM size and exclusively deep venous drainage. Initial mRS score a parts per thousand yen2, eloquent location and higher S-M grade may be associated with worsening functional status. Microsurgical resection can be safe and effective for selected patients. Preoperative embolization is helpful in patients with S-M grade IV-V AVMs. For those with surgical contraindications, SRS and observation are treatment alternatives.