The application of ultrasound to clinical evaluation of the symptomatic patient requires less precision and accuracy than when it is used as an instrument to study atherogenesis. The clinician's goal is to locate hemodynamically significant lesions; studies of atherogenesis require auxometry (measurement of the growth of lesions) and the instruments used must have a sensitivity matched to an average growth rate that can be expected in a general population. This dictates that ultrasound equipment be designed specifically for high resolution imaging and the target vessels of choice appear to be the carotid arteries. Correlations of disease patterns among various vessel beds indicate that inferences can be made between coronary and carotid vessels on a population level, but should not be made on an individual basis. While the ultimate goal of the cardiologist is to be able to do repetitive, accurate, noninvasive scanning of coronary arterial lesions, present technology limits us to making such inferences about coronary disease from the best available substitute. The carotid appears to be the artery of choice for this purpose because of its proclivity toward developing the disease and its ready availability for clinical scanning. It is an important target in its own right, and should be of interest to cardiologists given the likelihood that the coronary bypass patient will tend to develop atherosclerosis in the cerebral system. The ideal noninvasive imaging equipment for today's clinician is one which offers flexibility in being able to image various vessel beds with the same transducer head. This flexibility, however, may not be appropriate for equipment which is designed for auxometric application. Rather, lesion tracking image systems will need to be designed for narrow application: all variables will be controlled to maximize precise visualization of the carotid artery. If we were to project an ideal, hypothetical carotid scanning device to track lesion change over one year, it would probably have the following characteristics: It would be a B-mode imaging system rather than a Doppler in favor of ther increased axial resolution offered by B-mode. It would function at 8 MHz to 10 MHz to provide maximum axial resolution. The crystal would be the maximum size permitted by anatomy. The near/far field transition would be focused at the depth of the carotid (1 cm to 2 cm) to offer maximum azimuthal resolution. The system would utilize a head positioner to orient the patient's neck the same way on repeat scans.(ABSTRACT TRUNCATED AT 250 WORDS)