Measurement of the ST integral has usually incorporated ST-segment depression integrated between the J point and 80 ms after the J point (J + 80). To assess the effect of varied onset and offset of ST measurement on performance of the ST integral for the identification of three-vessel coronary disease, the exercise electrocardiograms (ECGs) of 60 patients with angiographically proven coronary disease were analyzed using J point or J + 20 onsets and J + 60 or J + 80 offsets of ST integral calculation. Simple ST-segment depression of greater than 200 microV, measured at 60 ms after the J point, identified three-vessel disease with a specificity of 68% (17 out of 25 patients) and sensitivity of 69% (22 out of 35 patients). At a matched specificity of 68% (17 out of 25 patients), there was identical sensitivity (54%, 19 out of 35 patients) of ST integrals measured either from the J point to J + 80, from the J point to J + 60, or from J + 20 to J + 60. A trend toward increased sensitivity (60%, 21 out of 35 patients) when the ST integral was measured from J + 20 to J + 80 did not reach statistical significance, and comparison of receive operating characteristics (ROC) curves demonstrated that varying the onset and offset of ST-segment measurement had no significant effect on the overall performance of ST integral criteria for the detection of three-vessel disease.(ABSTRACT TRUNCATED AT 250 WORDS)