This study was performed to determine if QT prolongation before and during early exercise is related to the occurrence of exercise-induced ventricular arrhythmias (EIVA). EIVA occurred in 47 of 142 patients with angiographic evidence of coronary artery disease (CAD); no EIVA occurred among the 22 patients without CAD (OV). Resting QTc and QT intervals during early exercise were similar in patients without EIVA, irrespective of the presence or absence of CAD; however resting QTc was significantly longer in CAD patients who showed EIVA (443 +/- 40 ms; P less than 0.01) than in CAD patients without EIVA (424 +/- 37 ms) and in OV patients (421 +/- 32 ms). During early exercise, the QT interval remained significantly longer in patients with than in those without EIVA. There was a trend toward increasing resting QTc in patients who exhibited EIVA more severe than grade 3. When resting QTc was longer than 440 ms, subsequent EIVA were correctly predicted in CAD patients with a sensitivity of 43%, a specificity of 72% and a predictive accuracy of 63%. Thus, a trend toward longer resting QTc values exists in CAD patients who develop EIVA; however, a long resting QTc (greater than 440 ms) appears to be only a weak predictor of subsequent EIVA.