This study was designed to establish the neurally-mediated effects of baroceptor and cardiopulmonary reflexes on heart rate in normal subjects. We therefore studied the effects of various interventions, able to modify the activity of arterial baroreceptors and cardiopulmonary receptors, on the power spectrum of the R-R interval in nine healthy men. To confirm the efficacy of these interventions left ventricular volume was monitored using a portable radionuclide probe. Isosorbide dinitrate (5 mgm sublingually) while sitting unloads both baroreceptors and the cardiopulmonary receptors; it decreased left ventricular end-diastolic volume (p < 0.01) and the mean R-R interval (p < 0.005), increased the power of the low-frequency (LF) component of the R-R interval (p < 0.01), decreased the power of the high-frequency (HF) component (p < 0.005) and increased the LF/HF ratio (p < 0.05). The application of cuffs around the thighs in the supine position, which unloads only cardiopulmonary receptors, decreased left ventricular end-diastolic volume (p < 0.05) but did not affect mean R-R interval, LF component, HF component and LF/HF ratio (p > or = 0.10). Leg raising, which loads only cardiopulmonary receptors, increased left ventricular end-diastolic volume (p < 0.05) and did not affect the mean R-R interval, LF component, HF component and LF/HF ratio (p > or = 0.10). In conclusion, our data suggest that, in normal man, baroreceptor unloading increased sympathetic and decreased parasympathetic neural control of heart rate; cardiopulmonary reflexes do not appear to have a direct neurally mediated effect on heart rate.