In humans, mu- and kappa-opioid receptor agonists lower tuberoinfundibular dopamine, which tonically inhibits prolactin release. Serum prolactin is, therefore, a useful biomarker for tuberoinfundibular dopamine. The current study evaluated the unexpected finding that the relative mu- and kappa-opioid receptor selective antagonist nalmefene increases serum prolactin, indicating possible kappa-opioid receptor agonist activity. In all, 33 healthy human volunteers (14 female) with no history of psychiatric or substance use disorders received placebo, nalmefene 3 mg, and nalmefene 10 mg in a double-blind manner. Drugs were administered between 0900 and 1000 on separate days via 2-min intravenous infusion. Serial blood specimens were analyzed for serum levels of prolactin. Additional in vitro studies of nalmefene binding to cloned human kappa-opioid receptors transfected into Chinese hamster ovary cells were performed. Compared to placebo, both doses of nalmefene caused significant elevations in serum prolactin (p<0.002 for nalmefene 3 mg and p<0.0005 for nalmefene 10 mg). There was no difference in prolactin response between the 3 and 10 mg doses. Binding assays confirmed nalmefene's affinity at kappa-opioid receptors and antagonism of mu-opioid receptors. [(35)S]GTPgammaS binding studies demonstrated that nalmefene is a full antagonist at mu-opioid receptors and has partial agonist properties at kappa-opioid receptors. Elevations in serum prolactin following nalmefene are consistent with this partial agonist effect at kappa-opioid receptors. As kappa-opioid receptor activation can lower dopamine in brain regions important to the persistence of alcohol and cocaine dependence, the partial kappa agonist effect of nalmefene may enhance its therapeutic efficacy in selected addictive diseases.