The study was aimed determine the criteria for treating cerebral aneurysms, ruptured and unruptured, suitable for endovascular Guglielmi detachable coiling (GDC) with least morbidity and mortality. We will discuss the importance of knowledge of the geometry of cerebral aneurysm and its impact on the results of coiling. We have treated 324 patients with cerebral aneurysms at the University of Illinois Hospital from May 1, 1994 to June 1997. During this period 139 patients were treated with GDC and 185 patients were treated surgically. Of 139 patients treated with coils 54 patients with 54 aneurysms were ruptured and the remaining 85 patients with 90 aneurysms were unruptured. For initial 25 patients, the selection of aneurysms for coiling was random and the geometry of the aneurysm was not considered as an important factor in the selection for coiling, however, in the later series of 114 patients with 119 aneurysms were selected for coiling only when the geometry looked favorable in the angiogram. With time we realized that the dome/neck ratio more than or equal to 2/1 and the diameter of the neck not exceeding 5 mm were most suitable for coiling. The initial series of 25 patients (May 1994 to February 1995) treated without taking geometry of the aneurysms as an important criteria for coiling led to high morbidity and mortality and less than 50% of these aneurysms were angiographically occluded at 6-month follow-up. In the second series of 114 patients (March 1995 to June 1997) with 119 aneurysms we had 0% mortality related to the coiling and only 1.8% permanent morbidity. We found complete aneurysm occlusion in 78% of the subarachnoid hemorrhage and 76% of unruptured group when the dome/neck ratio was more than or equal to 2/1. However, the occlusion rate dropped to 50% when the dome/neck ratio was less than 2/1. This preliminary experience suggests that GDC is a safe technique with low mortality-morbidity for the treatment of intracranial aneurysms in appropriately selected patients. The percentage of complete occlusion of the aneurysm following tight and dense packing is strongly dependent on the geometry of the aneurysm and we conclude that the best results are achieved when the dome/neck ratio is more than or equal to 2/1.