To determine whether balloon angioplasty can provoke arterial vasoconstriction independent of platelet aggregation and neurogenic input, we studied the spontaneous vasomotor effects of balloon dilatation in isolated, perfused whole-vessel segments of rabbit aorta and pig carotid artery. Freshly dissected rabbit thoracic aortas were mounted in a muscle bath-perfusion chamber, perfused with physiologic saline solution at 70 mm Hg, and allowed to equilibrate. The proximal or distal half of the aortas were dilated with either a "large" (5 mm, 31-51% stretch beyond relaxed diameter) or a "small" (4 mm, 5-16% stretch) balloon angioplasty catheter with the other half of the vessel serving as the control. A similar series of experiments were performed in pig carotid arteries using "large" (6 or 8 mm, 48-90% stretch) balloon catheters. The spontaneous vasomotor effects of balloon angioplasty were examined with long-axis, high-frequency ultrasonic imaging combined with computerized edge detection image processing to measure changes in segmental internal vessel diameters. Additional experiments were carried out in rabbit aortas to determine the roles of the endothelium, extracellular calcium, indomethacin, ibuprofen, and calcium-channel blockade in modulating angioplasty-induced vasoconstriction. Significant arterial vasoconstriction was observed in the balloon angioplasty segments after dilatation with 5-mm balloons but not with 4-mm balloons. After dilatation with 5-mm balloons, the angioplasty segments' cross-sectional areas decreased by an average of 31% versus 4% for the nondilated (control) segments (p less than 0.0001). Similar postangioplasty vasoconstriction was observed in the pig carotid arteries (decrease in minimal vessel cross-sectional area of 41% [angioplasty segment] versus 2% [control segment]) (p less than 0.005). This angioplasty-induced vasoconstriction was prevented by endothelial denudation before angioplasty, removal of extracellular calcium, and pretreatment with indomethacin or ibuprofen. The vasoconstriction was only partially inhibited by calcium channel blockade with verapamil. These findings demonstrate that stretch-pressure-induced arterial vasoconstriction may occur after balloon angioplasty, independent of platelet aggregation and neurogenic input. This angioplasty-induced vasoconstriction appears to be mediated by an endothelially derived cyclooxygenase product(s).