An attempt was made to assess the mechanism of directional coronary atherectomy using different methods of analysis. Quantitative coronary angiography was used as the gold standard to assess the immediate results of atherectomy, and a comparative quantitative analysis of atherectomy and balloon angioplasty was made. To determine whether the post-atherectomy cross-sectional area is close to a circle, we compared the area measurements obtained by edge detection with those obtained by videodensitometry. Finally, the extent of a 'Dotter' effect was established by quantitative angiography following crossing the stenosis with the atherectomy device. For the purpose of this study, the results of the first 113 successful atherectomy procedures were reviewed. In matched lesions, directional atherectomy induced a greater increase in minimal luminal diameter than balloon angioplasty (1.6 mm vs 0.8 mm; P < 0.0001). However, this luminal improvement is due to a substantial 'Dotter' effect induced by the bulky atherectomy device. Following atherectomy, only a slight difference in cross-sectional area measurements between edge detection and videodensitometry (mean difference: 0.28 mm2) was found. Histologic examination of an atherectomized coronary artery showed a near-circular postatherectomy area geometry. In conclusion, directional atherectomy is a very effective device with a substantially better initial result than balloon angioplasty. However, insertion of this bulky device itself causes an important 'Dotter' effect.