Cerebral blood flow increases upon the transition from rest to moderate exercise, but becomes affected when the ability to raise CO (cardiac output) is limited. HR (heart rate) is considered to contribute significantly to the increase in CO in the early stages of dynamic exercise. The aim of the present study was to test whether manipulation of the HR response in patients dependent on permanent rate-responsive ventricular pacing contributes to the increase in CO, MCA V(mean) [mean MCA (middle cerebral artery) velocity] and work capacity during exercise. The effect of setting the pacemaker to DSS ('default' sensor setting) compared with OSS ('optimized' sensor setting) on blood pressure, CO, SV (stroke volume) and MCA V(mean) was evaluated during ergometry cycling. From rest to exercise at 75 W, the rise in HR in OSS [from 73 (65-87) to 116 (73-152) beats/min; P<0.05] compared with DSS [70 (60-76) to 97 (67-117) beats/min; P<0.05] was larger. There was an increase in SV during exercise with DSS, but not with OSS, such that, at all workloads, SVs were greater during DSS than OSS. The slope of the HR-CO relationship was larger with DSS than OSS (P<0.05). From rest to exercise, MCA V(sys) (systolic MCA velocity) increased in OSS and DSS, and MCA V(dias) (diastolic MCA velocity) was reduced with DSS. No changes were observed in MCA V(mean). Manipulation of the pacemaker setting had no effect on the maximal workload [133 (100-225) W in OSS compared with 129 (75-200) W in DSS]. The results indicate that, in pacemaker-dependent subjects with complete heart block and preserved myocardial function, enhancing the HR response to exercise neither augments CO by a proportional offset of the exercise-induced increase in SV nor improves cerebral perfusion.