Various computerized methods with multiple parameter options for measurements of the QT interval now are available. The optimum parameter setting for most algorithms is not known. This study evaluated the influence of the threshold level applied on the T wave differential on the QT interval and its dispersion measured in normal and abnormal electrocardiograms (ECGs). Seven hundred sixty ECGs recorded in 76 normal subjects and 630 in 63 patients with hypertrophic cardiomyopathy (HCM) (10 consecutive recordings in each individual) were analyzed. In each lead of each ECG, the QT interval was measured by the threshold method applied to the first differential of the T wave. The threshold level was varied between 5% and 30% of the T wave maximum in 1% steps, resulting in 26 different choices of QT measurements. With each choice the maximum QTc and the QT dispersion (QTd, standard deviation of the QT in all 12 leads) were obtained for each recording. The maximum QTc was significantly longer in HCM patients than in normal subjects (P < 0.001) at all threshold levels except between 5% and 7%. The QTd was significantly greater in HCM patients at all threshold levels. The QTc and QTd changed significantly with the threshold level. The maximum QTc varied up to 60 ms in normal subjects and up to 70 ms in HCM patients, depending on the threshold level. Thus, the QT interval and its dispersion measured with the threshold method applied to the first T wave differential depended significantly on the threshold level in both normal and diseased hearts. All programmable options of available automatic instruments should be examined carefully before any study, and all algorithmic details should by systematically presented.