Neuromuscular blocking drugs exhibit different degrees of fade in response to train-of-four stimulation believed to represent their relative prejunctional effects. The present study was designed to compare the train-of-four fade after cisatracurium and compare this with other commonly used muscle relaxants. Train-of-four fade during onset and recovery of block were recorded after administration of cisatracurium 0.05 or 0.1 mg.kg-1, atracurium 0.5 mg.kg-1, vecuronium 0.08 mg.kg-1, mivacurium 0.15 mg.kg-1 or rocuronium 0.6 mg.kg-1 to patients anaesthetised with fentanyl, nitrous oxide and a propofol infusion. Neuromuscular monitoring was by stimulation of the ulnar nerve and recording the force of contraction of the adductor pollicis muscle. The onset and recovery of block were also measured. Train-of-four fade during onset of block was greater with the lower dose of cisatracurium compared with the higher dose of cisatracurium and all other relaxants. Train-of-four fade during recovery was similar. The median times (and ranges) for the onset of maximum block were 3.4 (2.1-5.6), 1.5 (1.2-2.3), 2.1 (1.2-2.6), 2.0 (1.5-2.7) and 1.0 (0.7-1.3) min for cisatracurium 0.1 mg.kg-1 and atracurium, mivacurium, vecuronium and rocuronium, respectively. The median times (and ranges) for the recovery of T1 to 25% of control and to a train-of-four ratio of 0.8 were 41 (21-50) and 65 (40-78); 43 (37-54) and 69 (58-79); 15 (11-20) and 25 (19-30); 31 (23-46) and 60 (45-117); and 33 (18-57) and 50 (28-76) min following cisatracurium, 0.1 mg.kg-1, atracurium, mivacurium, vecuronium and recuronium, respectively.