The pharmacokinetics and the hormonal, analgesic, and behavioral effects of several doses of human beta-endorphin were evaluated after intravenous administration to three patients and intracerebroventricular administration to one patient with pain caused by cancer. These effects were compared to the hormonal effects of intravenously administered morphine sulfate in two patients and an enkephalin analog in two baboons. The mean terminal half-life after intravenous administration of 5 or 10 mg of human beta-endorphin to three patients was 37 min; the mean volume of distribution was 178 ml/kg, and the metabolic clearance rate was 3.2 (ml/min)/kg. The half-life of beta-endorphin in cerebrospinal fluid after intracerebroventricular administration was 93 min, and the volume of distribution was 0.74 ml/kg. A rapid rise in plasma prolactin followed both intravenous and intracerebroventricular beta-endorphin. Intravenous administration did not affect plasma growth hormone, but intracerebroventricular administration suppressed plasma growth hormone. No significant change in plasma growth hormone was noted after intravenous administration of morphine to humans, but plasma growth hormone decreased in one baboon after administration of the enkephalin analog. beta-Endorphin-stimulated release of prolactin occurred at doses lower than those required to produce analgesic and other behavioral effects. When both hormonal and analgesic effects were observed (after 7.5 mg were given intracerebroventricularly), the onset of the hormonal response slightly preceded the analgesic and behavioral responses. These studies suggest that the hormonal effects of beta-endorphin are species dependent and are similar to those of morphine. Hormonal and analgesic effects of beta-endorphin appear to result from the activation of opiate receptors that differ in their locations and characteristics.