The purpose of this study was to quantify nonuniform rotation in a current mechanical intravascular ultrasound (IVUS) instrument and its effect on arc, area, and diameter measurements. The accurate reconstruction of IVUS two-dimensional images is dependent on uniform rotation of the catheter tip. Prior investigations suggested that bends in the catheter driveshaft may be responsible for poor torque transmission, nonuniform rotation, and consequent errors in IVUS measurements. Eight 30 MHz mechanically driven IVUS catheters were evaluated in a model simulating the catheter course through the aorta and coronary ostium in a clinical study. Angular velocity and posi-ion profiles of the transducer, image angle, and diameter and area measurement errors were obtained from each catheter by imaging a vascular phantom with eight equispaced echogenic markers from concentric and eccentric positions. Six catheters also were tested for comparison in a simple curvature model. Rotational error was found in all catheters tested and worsened in the aortic model. Maximal angular error, defined as the largest angle between actual and presumed transducer direction, increased when measured in the aortic model as compared with the simple curvature model (17 +/- 12 degrees to 45 +/- 25 degrees; p < 0.05). Angles of 45 degrees were misrepresented with a mean range of values of 26 to 63 degrees. With eccentric catheter placement, area and diameters had average maximal absolute errors of 26% +/- 7.8% and 23% +/- 10%, respectively. In conclusion, nonuniform rotation of mechanical IVUS transducers constitutes a significant potential source of error in IVUS measurement of arcs of calcification, and lumen shape, area, and diameter.