Indirect systolic time intervals (STI) corrected for heart rate were compared at rest and immediately following 3 minutes of moderate supine exercise in 33 normal subjects and 32 age- and sex-matched patients with coronary artery disease (CAD). The intervals were correlated with measurements of cardiac index (CI), stroke volume index (SVI), mean pulmonary artery and wedge pressures obtained under identical conditions of rest and exercise, and with resting left ventricular end-diastolic pressure (LVEDP), LV dp/dt, exercise factor, ejection fraction (EF), and extent of CAD.Total electromechanical systole (Q-A2c) was the same at rest in both normal and CAD groups and did not change with exercise in either group. The CAD group had a significantly longer preejection period PEPc), shorter left ventricular ejection time (LVETc), and larger PEP/LVET at rest and exercise than the normal group. Both groups responded to exercise with a significant reduction of PEPc, prolongation of LVETc, and reduction of PEP/LVET. Significant correlations were found between these STI and SVI, CI, LV dp/dt, LVEDP, and EF, which explained the differences in STI between the normal and CAD groups. However, exercise did not improve the sensitivities of the STI in detecting disordered hemodynamics in the patients with CAD. Furthermore, the STI failed to predict reliably hemodynamic abnormalities in the individual patient which were not already clinically obvious.