The rate of appearance of the sympathetic neurotransmitter norepinephrine (NE) in the regional venous drainage (NE spillover) can be estimated based on intravenous or intra-arterial infusions of [H-3]NE. The present study examined whether forearm NE spillover (FASO) in humans depends on forearm blood flow (FBF) and on the site of infusion of the tracer. Healthy volunteers underwent infusions of [H-e]NE and [H-3]isoproterenol (Iso) administered intravenously (n = 21), intra-arterially (n = 32), or by both routes in the same experimental session (n = 7). FBF was manipulated by intra-arterial infusions of the vasodilator sodium nitroprusside (n = 7) or the vasoconstrictor methoxamine (n = 7). Forearm extraction percents of [H-3]NE exceeded those of [H-e]Iso in all subjects undergoing intravenous infusions (54 vs. 46%, P < 0.001), whereas extraction percents of [H-3]Iso exceeded those of [H-3]NE when the tracers were infused intra-arterially. Regardless of the infusion site, FASO was positively correlated with FBF (r = 0.44, P < 0.005). Nitroprusside increased FBF and FASO, and methoxamine decreased FBF and FASO. When the tracers were added to whole blood, 89% of [H-3]NE and 83% of [H-3]Iso remained in plasma after 1 min; although no further loss of [H-3]NE occurred over time, only 60% of the added [H-3]Iso remained in plasma by 20 min. The results indicate that regional NE spillover is flow dependent, complicating inferences about regional sympathoneural activity. Intra-arterial infusion of [H-3]NE results in higher estimates of regional NE spillover than does intravenous infusion of the tracer. [H-3]NE and [H-3]Iso bind rapidly and differently to nonplasma components of blood, invalidating the use of the dual tracer approach to estimate regional neuronal uptake of NE when the tracers are administered intra-arterially.