OBJECTIVE: Fifteen basilar aneurysms approached via Kawase's anterior petrosectomy were analyzed to determine parameters that could reliably predict the applicability of this approach to specific basilar aneurysms on the basis of existing imaging. METHODS: Anatomic data were gathered by studying 40 dry skulls in which measurements were taken to define the limits of the surgical window. Clinical data were obtained from the review of charts and radiographic images of 15 patients surgically treated with the Kawase approach. The data were combined to categorize basilar aneurysms according to their position in relation to bony anatomy as seen on preoperative angiograms. RESULTS: Two relevant measurements were determined on lateral angiograms that were predictive of the applicability of operative approach. The K1 line determined the caudal extent of exposure of the Kawase approach to be 18 mm below the floor of the sella turcica and represented the distance to the floor of the internal auditory meatus. The K2 line determined the caudal extent of exposure of the posterior petrosectomy approach to be 24 mm below the floor of the sella turcica and represented the distance to the upper aspect of the jugular tubercle. Basilar aneurysms below the posterior clinoid process could be categorized in relationship to the regional bony anatomy in a manner that is predictive of the appropriate surgical approach as 1) retrosellar, 2) upper clival, 3) midclival, and 4) lower clival. Glasgow outcome data in 15 patients surgically treated with the Kawase approach demonstrated results comparable to those reported for ruptured basilar aneurysms. CONCLUSION: Individual basilar artery aneurysms can be categorized according to their relationship to bony anatomy on lateral view preoperative angiograms without bone subtraction. Anatomic parameters, the K1 and K2 lines, from these angiograms enable the neurosurgeon to predict the most appropriate approach for each type of basilar artery aneurysm.