The object of this study was to investigate whether pretreatment with pancuronium before i.v. injection of suxamethonium could cause prolonged neuromuscular blockade in patients heterozygous for the usual and the atypical plasma cholinesterase gene (E1uE1a). Forty-three patients, 23 with genotype E1uE1a and 20 with normal genotype (E1uE1u), were pretreated with pancuronium 0.01 mg.kg-1 followed by suxamethonium 1.5 mg.kg-1, and received either neurolept anaesthesia or halothane anaesthesia. Seven patients (E1uE1a) were given suxamethonium 1.5 mg.kg-1 without pretreatment. The duration and type of neuromuscular block were evaluated using train-of-four (TOF) nerve stimulation. Type of anaesthesia did not significantly influence the results. The duration of block following pretreatment was significantly longer in heterozygous patients than in normal patients. Time to 90% twitch height recovery was 10.7 +/- 1.2 min (mean +/- s.d.) in genotypically normal patients, and 18.0 +/- 4.2 min in patients with genotype E1uE1a. Pretreatment with pancuronium caused a significantly slower recovery of the TOF ratio (phase II block). Thus, a TOF ratio of 0.7 was always reached within 13 min in genotypically normal patients. In genotypically abnormal patients, the same TOF ratio was reached within 20 min in all but three patients. In these three patients time to 90% twitch height recovery was prolonged (18-31 min), and TOF ratio did not return to normal, but stabilized at about 0.35, 0.50, and 0.65, respectively. Injection of edrophonium restored normal neuromuscular function in 10 min. It is concluded that in patients heterozygous for the usual and the atypical gene, pretreatment with pancuronium in combination with an increased dose of suxamethonium may cause a phase II block and thus a prolonged neuromuscular block.