Automation of blood pressure (BP) measurements during exercise has proved difficult because motion artifact is a major limitation in ultrasound techniques and noise artifact limits the application of sound transcution via microphone pickups. We assessed the value of a new automated system of indirect BP determination by comparing it with manually determined BP in 50 consecutive patients referred for treadmill testing. Automated BP determinations were blinded to the physician or technician who was simultaneously manually auscultating BP. The automated system uses acoustic transduction, but with ECG assist and microprocessing of nonsynchronous noise to overcome the limitations of other systems. The data were statistically analyzed and the correlation coefficient, mean difference and standard deviation of the difference for systolic and diastolic BP for differing levels of heart rate and treadmill stage were determined. The correlation between automated and manually determined BP was 0.964 for systolic BP and 0.848 for diastolic BP. Despite some limitations, this automated system offers significant advantages for exercise BP determination.