New coronary devices are being developed in attempts to solve the limitations of balloon angioplasty. Directional coronary atherectomy was the first of these devices to gain Food and Drug Administration approval. Theoretically, directional atherectomy improves coronary stenoses by an entirely different mechanism than balloon angioplasty, removing atherosclerotic plaque and leaving a smooth surface with less elastic recoil. Nonrandomized experience has shown that directional atherectomy is most useful in proximal to mid-segment, noncalcified, large (> 2.5 mm) coronary arterial segments, yielding minimal residual stenosis and larger lumen diameters than are generally achieved with angioplasty. Experience suggests that directional atherectomy is most useful in ostial coronary lesions, bifurcation stenoses, proximal left anterior descending lesions, discrete saphenous vein graft stenoses, complex or thrombus-containing lesions, highly eccentric lesions, and lesions failing PTCA secondary to elastic recoil, recurrent thrombosis, or limited dissection. Dissection and out-of-catheterization-laboratory acute closure are infrequent and appear to be improved over balloon angioplasty. However, intimal hyperplasia leading to restenosis is not solved by directional atherectomy. Randomized studies such as the Coronary Angioplasty Versus Excisional Atherectomy Trial (CAVEAT) will help to clarify further the overall utility of directional atherectomy compared with balloon angioplasty in the future, but due to the limited power of subgroup analyses, CAVEAT may not have fully defined all the specific lesion characteristics that may be improved by atherectomy over PTCA. Review of the data presented may help the clinician identify specific areas in which atherectomy appears to offer an advantage over angioplasty, although definitive answers await specifically targeted randomized trials.