OBJECTIVE Our goal was to determine the accuracy of MR angiography at 0.5 T for the diagnosis of intracranial aneurysms. METHODS We retrospectively studied 140 patients, 70 with acute subarachnoid hemorrhage, who were either at high or low risk for intracranial aneurysm. Three-dimensional time-of-flight MR angiography was typically performed to cover the circle of Willis, with a volume thickness of 30 mm. Conventional spin-echo MR images and MR angiograms were reviewed together, and the results were compared with those obtained at intraarterial cerebral angiography to determine the sensitivity and specificity of MR angiography. RESULTS Eighty-nine aneurysms (size range, 2 to 27 mm; 25 aneurysms < 5 mm) were identified at intraarterial cerebral angiography. Six aneurysms were missed by MR angiography and two were doubtful (sensitivity, 91% to 93%; specificity, 100%). Missed aneurysms were located outside the MR angiographic acquisition volume (n = 3) or on the carotid siphon (n = 3; size = 2, 3, and 5 mm). CONCLUSIONS Even if MR angiography presents some restrictions in acquisition volume and spatial resolution, the detection rate of intracranial aneurysms is excellent at 0.5 T in both asymptomatic patients and in those with subarachnoid hemorrhage. A midfield system is not a restriction to the detection of intracranial aneurysms by MR examination.