Accurate detection and measurement of the P-wave and T-wave components of the ECG complex have been difficult and often avoided in computerized ECG analysis. This is particularly so during exercise testing where T-P fusion occurs during higher heart-rates. By combining advanced pattern-recognition techniques, statistical measurements and empirically based heuristic decision-making logic, our ECG Analysis program has been able to reliably detect, measure and track these components during exercise testing to a degree surpassing visual detection by highly experienced readers. Complete analysis of each consecutive record gathered during the exercise test is performed and a data-base of measurements and parameters is created for reference comparison of previous results at each analysis step to current measurements. Thus, evaluation of each current record for appropriate and accurate analysis is based on an expert system approach which is constantly updated and can adjust itself to individual ECG morphology as the test progresses. Preliminary attempts are also being made to separate, extract and normalize the P-wave and T-wave during fusion for better understanding and comparison of changes which occur at high heart-rates. Theoretical and clinical reasons related to the detection and measurements of the P-wave and T-wave during exercise testing are also discussed.