In both normotensive and hypertensive subjects blood pressure (BP) and heart rate (HR) show concordant changes over 24 h. This may depend on a central factor exerting influences of the same nature on cardiac and vascular targets. An alternative explanation, however, is that a cause-effect relationship links these variabilities, i.e. that HR variations induce BP changes [presumably via variations in cardiac output (CO)]. Blood pressure was recorded intra-arterially in five supine and five exercising (walking) essential hypertensive subjects during a control period of 1 h and during an additional hour in which atropine, 0.04 mg/kg body weight, was injected intravenously (i.v.). The same recordings were performed in seven other subjects, in which saline rather than atropine was employed. One-hour BP and HR variabilities (variation coefficients, VC) were computer analysed. In both the supine and the exercising subjects atropine caused a marked reduction in HR VC (-65.3 and -48.4%, respectively). In the supine subjects this reduction was accompanied by only a modest reduction in BP VC whereas in the exercising subjects the BP VC increased by 30.4%. In the seven subjects in which saline was injected no change in BP and HR VC occurred. Thus a marked reduction in HR variability is not accompanied by a comparable attenuation in BP variability, which rules out a cause-effect link between these two phenomena. Indeed, during physical exercise HR stabilization is followed by an increase rather than a reduction in BP variation, which supports the conclusion that under some circumstances HR plays an anti-oscillatory role.