In dogs, manipulation of heart rate has no effect on the exercise-induced increase in cardiac output. Whether these findings apply to humans remain uncertain, because of the large differences in cardiovascular anatomy and regulation. To investigate the role of heart rate and peripheral vasodilatation in the regulation of cardiac output during steady-state exercise, we measured central and peripheral haemodynamics in 10 healthy male subjects, with and without atrial pacing (100–150 beats min(−1)) during: (i) resting conditions, (ii) one-legged knee extensor exercise (24 W) and (iii) femoral arterial ATP infusion at rest. Exercise and ATP infusion increased cardiac output, leg blood flow and vascular conductance (P < 0.05), whereas cerebral perfusion remained unchanged. During atrial pacing increasing heart rate by up to 54 beats min(−1), cardiac output did not change in any of the three conditions, because of a parallel decrease in stroke volume (P < 0.01). Atrial pacing increased mean arterial pressure (MAP) at rest and during ATP infusion (P < 0.05), whereas MAP remained unchanged during exercise. Atrial pacing lowered central venous pressure (P < 0.05) and pulmonary capillary wedge pressure (P < 0.05) in all conditions, whereas it did not affect pulmonary mean arterial pressure. Atrial pacing lowered the left ventricular contractility index (dP/dt) (P < 0.05) in all conditions and plasma noradrenaline levels at rest (P < 0.05), but not during exercise and ATP infusion. These results demonstrate that the elevated cardiac output during steady-state exercise is regulated by the increase in skeletal muscle blood flow and venous return to the heart, whereas the increase in heart rate appears to be secondary to the regulation of cardiac output.