BACKGROUND: The Pipeline Embolization Device (PED) has emerged as a promising treatment for intracranial aneurysms. OBJECTIVE: To assess the safety and efficacy of the PED, to analyze the effect of operator experience on the complication rate, and to identify predictors of complications and obliteration. METHODS: A total of 109 patients with 120 aneurysms were treated with PED at our institution. The patient population was divided into 3 consecutive equal groups to assess whether overall and major complication rates decreased over time: group 1, patients 1 through 37; group 2, patients 38 through 73; and group 3, patients 74 through 109. RESULTS: The number of PEDs used was 1.40 per aneurysm. Symptomatic and major procedure-related complications occurred in 11% and 3.7% of patients, respectively. The rate of complications decreased from 16.2% in group 1 to 5.6% in group 3, and the rate of major complications fell dramatically from 10.8% in group 1 to 0% in groups 2 and 3 (P < .05). Procedure time significantly decreased over time (P = .04). In multivariate analysis, previously treated aneurysms were predictive of procedural complications (P = .02). At the latest follow-up, 65.8% of aneurysms were completely occluded, 9.6% were nearly completely occluded, and 24.6% were incompletely occluded. In multivariate analysis, fusiform aneurysms (P = .05) and shorter angiographic follow-up (P = .03) were negative predictors of aneurysm obliteration. CONCLUSION: PED therapy may have an acceptable safety-efficacy profile. The risk of complications appears to decrease dramatically with physician experience, supporting the existence of a learning curve. Patients with previously treated aneurysms have higher complication rates, whereas fusiform aneurysms achieve lower obliteration rates.