Treatment of septic shock is a persistent dilemma. The clinical use of agents such as naloxone has resulted in variable success. Because the dosage and timing of these agents are considered critical factors in their efficacy, we investigated both dosage and timing of naloxone. Thirteen consecutive patients with documented septic shock and resistance to a one-liter fluid challenge underwent invasive hemodynamic monitoring and the administration of naloxone by initial bolus of 0.03 mg/kg followed by infusion at a rate of 0.2 mg/kg.h over one hour. During the one-hour observation period, iv fluid administration, concomitant pressor agents, and respirator values were constant. After infusion, adjustments in fluid administration, respirator status, and pressor agents were made as required by the clinical situation. A significant increase in mean arterial pressure (MAP) over baseline (60 +/- 3 mm Hg) was noted at 5 min (77 +/- 6 mm Hg, p less than .005) and at 30 min (73 +/- 6 mm Hg, p less than .025). Similarly, a significant increase in systolic arterial pressure was noted over prenaloxone levels (89 +/- 3 mm Hg) at 5 min (114 +/- 6 mm Hg, p less than .001), 30 min (107 +/- 8 mm Hg, p less than .05), and at one hour (106 +/- 8 mm Hg, p less than .05). There was a moderate nonsignificant increase in cardiac index, pulmonary capillary wedge pressure, and systemic vascular resistance. No side-effects to naloxone were noted in our group. No effect on survival could be demonstrated. We found no overall effect on mortality. However, by its increase of MAP, naloxone may serve as a temporizing agent during the treatment of critically ill patients with septic shock.