1. Although the opiate antagonist naloxone has been shown to affect sympathoadrenomedullary function in some studies, this has not been a uniform finding in all investigations, using different doses of naloxone. We have therefore investigated the actions of saline placebo and increasing bolus doses of intravenous naloxone (25, 100 and 250 μg/kg) on the plasma noradrenaline, adrenaline, adrenocorticotrophin (ACTH) and cortisol responses to a cold-pressor test in six males and two females in a double-blind randomized study. 2. Basal levels of adrenaline did not differ on any of the study occasions: the cold-pressor stimulus produced a significant rise in mean plasma adrenaline to a peak of 0.16 nmol/l, with a peak incremental rise of 0.08 nmol/l. In the presence of the two higher doses of naloxone, the incremental rise in the mean adrenaline level was significantly enhanced, reaching 0.30 nmol/l at 100 μg of naloxone/kg and 0.29 nmol/l at 250 μg of naloxone/kg, with no significant enhancement observed at the lowest dose of naloxone. The rise in plasma noradrenaline, systolic and diastolic blood pressure and pulse rate during the cold-pressor test was not consistently altered by any dose of naloxone, but there was a significant trend for the degree of discomfort to increase with the dose of naloxone. 3. Neither plasma ACTH nor serum cortisol rose in response to the cold-pressor stimulus. Analysis of variance demonstrated a significant naloxone-treatment effect for both ACTH and cortisol, with the 100 μg/kg and 250 μg/kg doses of naloxone leading to a significant rise in mean plasma ACTH not seen at the lowest dose of 25 μg/kg. 4. In a separate study of the effect of naloxone on the clearance of plasma adrenaline, adrenaline was infused on two occasions into eight normal male subjects to a maximum dose of 5 μg/min for 15 min, and blood was sampled at regular intervals during and after termination of the infusion, preceded by 8 mg of naloxone or placebo. After termination of the adrenaline infusion, there was a rapid fall in plasma adrenaline (Λ = −1.9 ± 0.3 min), which was not altered by naloxone (Λ = −1.8 ± 0.3 min). 5. We conclude that the potentiation of the adrenaline response to a cold-pressor test by naloxone is mediated by increased secretion of adrenaline and not by a reduction in its clearance rate, and that this response is of low naloxone sensitivity.