To test the hypothesis that tortuous, diseased vessels could be successfully treated with a flexible rotational atherectomy device we evaluated the BARD atherectomy device with quantitative angiography and histology in normal canine coronary arteries and diseased human below-knee amputation specimens. The mid left anterior descending and the circumflex vessels were treated in 4 dogs serially with 1 wk separating treatments. The acute and follow-up anterior descending artery size was unchanged (1.41 mm before, 1.39 mm after, and 1.59 mm at 1 week). Similar findings were obtained in the circumflex vessels. In 4 adult human amputated legs, diseased peroneal or tibial arteries were treated with a significant reduction in the percent luminal diameter stenosis from 62.75 prior to intervention to 36.5 following intervention (p = 0.021). The luminal diameter increased from 0.81 to 1.54 mm (p = 0.06). In 2 canine arteries there was histologic evidence of localized perforation of the arterial wall, but there was no angiographic evidence of perforation or dissection and no significant myocardial necrosis in the distribution of the treated vessels at 1 wk. The majority of the diseased human vessels demonstrated smoothly cut atheromas with sparing of the media. The rotational atherectomy catheter system holds promise for removal of plaque in relatively small, diffusely diseased, tortuous vessels.