In isolated organs, or when given in low dose intra-arterially, tricyclic antidepressant drugs are known to block reuptake of norepinephrine into sympathetic nerve varicosities, with a resultant increased norepinephrine washout. On the other hand, systemic administration of such drugs in humans reduces norepinephrine spillover to plasma. To clarify these seemingly contradictory findings, we have measured concurrently muscle sympathetic activity in the peroneal nerve (microneurography) and rates of norepinephrine spillover to plasma for the body as a whole and for the heart, the kidneys, and the forearm (radiotracer technique), both before and after intravenous infusion of desipramine, 0.5 mg/kg. Desipramine lowered the overflow of norepinephrine to plasma for the body as a whole and from the forearm and the kidneys (by 30-50%) but increased cardiac norepinephrine spillover by 25%. Both the number of sympathetic bursts per min in the peroneal nerve and their mean voltage amplitudes were markedly reduced after desipramine; total activity (bursts/min x mean burst amplitude) fell by approximately 90%. The effects of desipramine on norepinephrine spillover are explicable in terms of inhibition of central sympathetic outflow, balanced against the local blockade of transmitter reuptake. In most sites, the predominant effect is a reduction of norepinephrine overflow. For the heart, where reuptake is so important in transmitter disposition, the net effect is increased overflow.