Using a specific and sensitive radioimmunoassay, naloxone concentrations in the brains and sera of rats were measured at intervals for four hours following iv injection (5 mg/kg). Decrement curves of naloxone were compared with those after iv injection of morphine (5 mg/kg). Serum concentration of naloxone at 5 minutes was 1.45 +/- 0.1 mug/ml (mean +/- SE) and that of morphine was 1.0 +/- 0.08 mug/ml. Their serum half-lives from one to four hours were approximately the same, 30-40 minutes. With naloxone, the brain-serum concentration ratios ranged from 2.7 to 4.6. Concentration of naloxone in the brain declined parallel to that in the serum. However, with morphine the initial brain concentration was approximately one tenth that in the serum (0.096 +/- 0.04 mug/ml). The brain morphine concentration was sustained for one hour, while serum morphine concentrations declined from 1.0 to 0.19 mug/ml during this period. Two minutes after iv injection of naloxone HCl (0.4 mg) in nine healthy volunteers, the serum drug concentration was 0.01 +/- .001 mug/ml. At 5 minutes, 97 per cent of the administered dose was no longer found in the serum, the serum concentration being 0.004 +/- .0003 mug/ml. From 20 minutes to two hours after injection, the calculated mean serum half-life of naloxone was 64 minutes. These results suggest that the rapid penetrance of naloxone into the brain and the high brain-serum concentration ratio contribute to its rapid onset of action and potency as a narcotic antagonist. The rapid decline of naloxone concentration in the brain found in the animal model, in contrast to that of morphine, could be the basis for its relatively short duration of action.