We evaluated the human pharmacology of dynorphin A(1-13) and determined whether this peptide can modulate naloxone-precipitated withdrawal effects. Such information could help determine its receptor mechanism of action and whether dynorphin is useful for treating opioid dependence. Six opioid-experienced subjects participated in a within-subject, placebo-controlled design. There were two phases, each with four test sessions. In phase 1, volunteers who were not physically dependent were administered 0, 0.1, 0.32 and 1 mg/kg dynorphin (15-min i.v. infusion) in ascending order, and subjective, observer-rated and physiological effects were monitored. Dynorphin produced brief, dose-related increases in drug effect ratings with both good and bad drug effects reported by different subjects. There were no significant changes in pupil size, respiratory rate, skin temperature, heart rate or blood pressure, These data are consistent with preclinical findings that dynorphin has a short duration of action and does not primarily exert its direct effects through mu-opioid receptors. In four separate sessions of phase 2, acute morphine pretreatment (45 mg/70 kg i.m.) was followed 15 or 18 hr later by dynorphin (0 vs. 1 mg/kg, 15-min i.v.) and then naloxone (1 or 3 vs. 10 mg/70 kg, 5-min i.v.). Under these conditions, dynorphin weakly potentiated naloxone-precipitated withdrawal. These data contrast with those of previous preclinical studies showing dependence-attenuating effects of dynorphin and fail to support its use as an antiwithdrawal agent in humans.