Background and Purpose-Childhood intracerebral hemorrhage is mainly attributable to underlying brain arteriovenous malformations (bAVMs). Multimodal treatment options for bAVMs include microsurgery and embolization, allowing an immediate cure, and radiosurgery, entailing longer obliteration times. Follow-up data on pediatric ruptured bAVMs are scarce, making it difficult to assess the risk of subsequent intracerebral hemorrhage. Our aim was to assess the clinical and angiographic outcome and to analyze risk factors for rebleeding during and after combined treatment of pediatric bAVMs. Methods-A prospectively maintained database of children referred to our institution between January 1997 and October 2012 for bAVMs was retrospectively queried to identify all consecutive ruptured bAVMs treated by surgery, embolization, and radiosurgery. The impact of baseline clinical and bAVM characteristics on clinical outcome, rebleeding rate, annual bleeding rate, and bAVM obliteration was studied using univariate and multivariate Cox regression analysis. Results-One hundred six children with ruptured bAVMs were followed up for a total of 480.5 patient-years (mean, 4.5 years). Thirteen rebleeding events occurred, corresponding to an annual bleeding rate of 2.71±1.32%, significantly higher in the first year (3.88±1.39%) than thereafter (2.22±1.38%; P<0.001) and in the case of associated aneurysms (relative risk, 2.68; P=0.004) or any deep venous drainage (relative risk, 2.97; P=0.002), in univariate and multivariate analysis. Partial embolization was associated with a higher annual bleeding rate, whereas initial surgery for intracerebral hemorrhage evacuation was associated with a lower risk of rebleeding. Conclusions-Associated aneurysms and any deep venous drainage are independent risk factors for rebleeding in pediatric ruptured bAVMs. Immediate surgery or total embolization might be advantageous for children harboring such characteristics, whereas radiosurgery might be targeted at patients without such characteristics. (Stroke. 2014;45:1664-1671.)