Low-frequency (less than 0.15 Hz) fluctuations of heart rate are increased by maneuvers, such as standing or hemorrhage, that increase sympathetic outflow to the heart and vasculature. To test the hypothesis that low-frequency heart rate fluctuations provide an index of sympathetic efferent activity, we compared power spectral measures of heart rate variability with two measures of sympathetic outflow, peroneal nerve sympathetic activity and antecubital vein plasma norepinephrine concentrations. Autonomic outflow was varied with graded stepwise infusions of nitroprusside and phenylephrine, which lowered or raised average diastolic pressures by approximately 15 mmHg. Before vasoactive drug infusions, no spectral measure of heart rate variability correlated significantly with muscle sympathetic activity, plasma norepinephrine concentration, average heart rate, or arterial pressure. During increases of muscle sympathetic activity and probable reductions of cardiac vagal activity induced by nitroprusside, the fraction of heart rate spectral power at low frequencies, but not the absolute value, correlated significantly with muscle sympathetic activity and plasma norepinephrine. However, during reductions of muscle sympathetic activity and probable elevations of cardiac vagal activity induced by phenylephrine, no measure of heart rate variability correlated significantly with muscle sympathetic activity. These findings can be explained by a model of heart rate control in which low-frequency heart rate fluctuations result from changing levels of both the sympathetic and parasympathetic inputs to the sinoatrial node.