The first report of the International Study of Unruptured Intracranial Aneurysms (ISUIA) cited very low rates of subarachnoid hemorrhage (SAH) for previously unruptured aneurysms. 1 Data were collected retrospectively. It was difficult to reconcile these rates with those reported from prior studies, with the known sizes of ruptured aneurysms and with clinical experience. Prior reports including prospective studies suggested that unruptured aneurysms carried a 1% to 2% risk of hemorrhage per year. [2][3][4] The second report from the ISUIA prospectively enrolled 4060 patients with unruptured aneurysms. 5 Of these, 1692 had no intervention for their aneurysm, 1917 had open surgery, and 451 had endovascular procedures. The rupture rates were more in agreement with prior studies (Table). Risk factors for SAH were increasing aneurysm size and aneurysm location at the basilar apex or posterior communicating artery. The risk of surgical repair increased significantly with increasing patient age, posterior circulation location of the aneurysm, history of ischemic cerebrovascular disease, and presence of symptoms from the aneurysm. Thirty-day mortality occurred in 1.5%, morbidity in 3%, poor cognitive function plus morbidity (a Rankin score of 3 to 5) in 4%, and overall total morbidity and mortality in 13%. The data indicate that the decision to treat a ruptured aneurysm needs to include a careful analysis of the patient, their risk factors for poor outcome, and the features of the aneurysm. Endovascular treatment is an option, although overall complete obliteration rates were only about 50% in this study, and risk of treatment was similar to but lower than with surgery, although the groups were not randomized to treatment and are thus not comparable. The findings of the initial results of ISUIA that were at variance with prior assumptions about unruptured aneurysms prompted an exhaustive review of the literature that suggested some explanations for the findings. 6 Numerous assumptions and biases inherent in the retrospective design of the first study were cited. An important one is that selection bias could account for the finding that 10 mm seemed to be the cutoff for rupture, yet this is at odds with the fact that the average size of a ruptured aneurysm is 8 mm. The distribution of cases included in the initial cohort will strongly influence the size cutoff for rupture that is found. To take an extreme example, if one collects only aneurysms Ͼ10 mm in diameter, then one will necessarily conclude that the cutoff for rupture is some value Ͼ10 mm. The author recommended that unruptured aneurysms with clinical pathological profiles resembling those of ruptured lesions be considered for treatment at a smaller size than unruptured lesions with profiles typical of intact aneurysms. The question of whether and how to follow patients with unruptured aneurysms that are not treated is unanswered, but it was felt that periodic radiological imaging might be wise. Screening for aneurysms in asymptomatic patients is only recommended...