The selection and administration of neuromuscular blocking (NMB) drugs in intensive care unit (ICU) patients remain controversial. We compared the dose-response and recovery pharmacodynamics of a new intermediate-acting NMB drug, cisatracurium besylate, to the intermediate-acting NMB drug, vecuronium (VEC), in a prospective, randomized, double-blind, multicenter study in critically ill adults. After informed consent, 58 mechanically ventilated ICU patients from five medical centers were randomized to receive either cisatracurium or VEC. Fifty-four of the 58 patients received NMB drugs before entering this study but demonstrated at least partial recovery (> or = one twitch) in the train-of-four (TOF) response before initiation of the NMB study drug. NMB drug infusion was titrated by peripheral nerve stimulation to maintain at least one twitch in the TOF response. NMB drugs were infused for 1-5 days. After discontinuation of NMB drug infusion, recovery of neuromuscular transmission was monitored with an accelerometer. NMB drug infusion for 28 cisatracurium patients averaged 2.6 +/- 0.2 (mean +/- SEM) micrograms.kg-1.min-1 with a mean duration of 80 +/- 7 h. After discontinuing cisatracurium administration, recovery to 70% TOF ratio averaged 68 +/- 13 min. The mean infusion rate for 30 VEC patients was 0.9 +/- 0.1 micrograms.kg-1.min-1 with a mean duration of 66 +/- 12 h. Neuromuscular recovery after VEC averaged 387 +/- 163 min, which was significantly longer (P = 0.02) than that after cisatracurium. Prolonged recovery of neuromuscular function after discontinuation of NMB drug infusion (identified by the primary investigator at each medical center) was reported in two cisatracurium patients and 13 VEC patients (P = 0.002), and occurred despite the routine use of neuromuscular twitch monitoring. Seven VEC and one cisatracurium patients died during the infusion of study drug or within 48 h after discontinuation of the NMB drug infusion. In summary, we found recovery of neuromuscular function after discontinuation of NMB drug infusion in ICU patients is significantly faster with cisatracurium than with VEC. In addition, routine neuromuscular monitoring was not sufficient to eliminate prolonged recovery and myopathy in ICU patients.