T-wave morphology classification and recognition plays an important role in clinical diagnosis based on Electrocardiogram (ECG). The integrated method based on Principal Component Analysis (PCA), threshold, linear regression, and symbolic method was used to analyze T-wave morphologies in this paper, and it was simple and convenient to implement. All kinds of T-wave shapes, expressed by symbol 'ABCDE', could be included by this method, and the recognition results were obvious, which could provide a reliable scientific basis for the clinical detection. After European ST-T database verification, the accuracy of T-wave morphology analysis was above 92%.