The recommended treatment of human chloroquine poisoning is diazepam and adrenaline but neither has been evaluated in controlled clinical trials. We investigated whether diazepam provided any added benefit over barbiturate anaesthesia and whether the protective effect of catecholamines in chloroquine poisoning was mediated through alpha or beta receptor stimulation. Rats, anaesthetised with thiobutobarbitone had a continuous intravenous infusion of 3 mg/kg/min of chloroquine. This caused a steady decline in pulse rate and blood pressure. When diazepam (3 mg/kg iv) was administered 5-10 min later, heart rates decreased at a faster rate (P = 0.005), blood pressure was consistently lower (P = 0.01) and there was a shorter time to arrhythmias and death (P < 0.05). Adrenergic agents were given by titration to attempt to maintain mean blood pressure > 75 mmHg. Compared with the phenylephrine (selective alpha agonist) group, the group treated with isoprenaline (selective beta agonist) had faster heart rates which decreased more slowly (P < 0.0001), higher blood pressure (P = 0.005) and longer time to arrhythmias and death (P = 0.005). Adrenaline and noradrenaline had intermediate effects. Thus beta agonist effect appears to explain the beneficial effects of adrenaline but alpha agonist activity may be harmful. This animal work suggests that a combination of barbiturate anaesthesia and isoprenaline may be better than the diazepam and adrenaline in combatting the effects of chloroquine.