This study was performed to establish the accuracy of the Avionics 445B recorder and Innovator 750 scanner for identifying ST-segment changes in patients undergoing treadmill exercise testing. A series of 143 patients underwent treadmill exercise testing with 12-lead ECG monitoring and 2-lead Holter monitoring. The mean age of the patients (134 men and 9 women) was 59.3 years (range, 27 to 77). The frequency range of the recorder was 0.08 to 30 Hz. A modified V2 and V5 bipolar lead system was used for the Holter recording. The Holter tape was then scanned using the Innovator 750 scanner. The technician adjusted the baseline-reference point and J-point and then did not interact further with the recording. Treadmill exercise testing was undertaken using a submaximal Bruce protocol with 12 standard ECG leads recorded. Forty-two patients had positive ST-segment changes recorded by both the 12-lead ECG and the Holter monitor. Ninety-two patients had no ST-segment changes recorded by either of the systems. With the 12-lead ECG recorded during the treadmill exercise test used as the standard, the sensitivity of the Holter system was 89%, specificity 96%, and positive predictive accuracy 91%. The Holter recorder and automated scanning system should reliably detect episodes of ST-segment change occurring during ambulatory ECG monitoring.