This report describes a single site experience as part of a multicenter clinical trial with high-speed rotational atherectomy in human coronary arteries. A total of 108 patients with 143 lesions had interventions, were grouped by success or failure, and were analyzed by patient, lesion, and procedural variables. Satisfactory results were achieved in 131 of 143 lesions (92%) and 99 of 108 (92%) patients. Neither patient-related variables (age, gender, diabetes, hypertension, cigarette use, restenosis, previous myocardial infarction, and left ventricular function) nor lesion characteristics (length, ostial or bifurcation location, calcification, lesion classification, and coronary location) were predictive of poor outcome. Tears, acute closure, percentage stenosis after rotational atherectomy and after adjunctive balloon angioplasty were the procedural variables that were statistically associated with outcome by univariate methods. Multivariate analysis isolated postintervention residual stenosis as the only variable that was statistically different between groups. Serious complications included one death in the catheterization laboratory, one Q-wave myocardial infarction, three non-Q myocardial infarctions, and three emergency coronary bypass operations for sustained vessel closure. One patient required emergency surgery for a pacing wire perforation not related to the use of the device. The potential benefits of high-speed rotational atherectomy include increased safety in complex lesions, the ability to address lesions not amenable to balloon techniques, and the possibility of reducing the incidence of restenosis.