Unruptured cerebral aneurysms are commonly treated by surgical clipping, but endovascular coil embolization is increasingly employed as an alternative. In a blinded review of unruptured aneurysms treated at our institution since 1990, we identified patients whose aneurysms were judged to be treatable by both neurosurgeons and neurointerventional radiologists. A change in Rankin Scale score of 2 or more from hospital admission to discharge, indicating a new moderate disability or worse, was predefined as the primary outcome measure. Long-term follow-up was obtained by mailed questionnaire and telephone interview. Length of stay and hospital charges were totaled for all hospitalizations, including follow-up. Sixty-eight patients treated surgically and 62 patients treated with endovascular coil embolization were considered candidates for either procedure on blinded review, and overall anticipated procedure risk was rated as identical. A larger proportion of patients in the surgical group developed a change in Rankin Scale score of 2 or more (25% of surgical patients vs 8% of endovascular patients). Total length of stay was longer (mean days: 7.7 for surgical patients vs 5.0 for endovascular patients) and hospital charges were greater (mean, $38,000 for surgical patients vs $33,400 for endovascular patients) for the surgical patients. At follow-up, an average of 3.9 years after the procedure, surgical patients were more likely to report persistent new symptoms or disability since treatment (34% of surgical patients vs 8% of endovascular patients) and a longer period for recovery to normal (50% returning to normal in 1 year for surgery and in 27 days for coil embolization). Coil embolization of unruptured cerebral aneurysms seems to be associated with significantly fewer complications than surgical clipping. More long-term data on aneurysm rupture rates are required to confirm efficacy.