Objectives:To compare recovery times from neuromuscular blockade between two groups of critically ill patients in whom pancuronium was administered by continuous infusion or intermittent bolus injection. To compare the mean pancuronium requirements (milligrams per kilogram per hour) and to assess the incidence of prolonged recovery times (>12 hrs) and residual muscle weakness.Design:Prospective, observational cohort.Setting:Intensive care unit in a university-affiliated hospital.Patients:A total of 30 mechanically ventilated patients who required pharmacologic paralysis. Patients were excluded if they had renal failure (creatinine clearance <30 mL/min), heart rate >130 beats/min, hepatic failure, peripheral nerve disease or myopathy, stroke, spinal cord damage, or myasthenia gravis.Interventions:Patients were assigned to receive pancuronium either by continuous infusion (n = 14) or intermittent bolus (n = 16). Depth of paralysis was titrated to maintain one or two responses to Train-of-Four stimulation with an accelerograph and desired clinical goals. Recovery time was defined as time from discontinuation of muscle relaxant until the amplitude of the fourth twitch, measured every 15-30 min using an accelerograph, was 70% the amplitude of the first twitch (Train-of-Four ≥ 0.7).Measurements and Main Results:These patients included the only three patients with status asthmaticus in our study. The groups were similar with respect to age, sex, weight, Acute Physiology and Chronic Health Evaluation II score, mode of ventilation, creatinine clearance, indications for paralysis, and duration of pancuronium administration. The median time for patients to recover from paralysis was 3.5 hrs (95% confidence interval, 1.82-5.18) in the infusion group vs. 6.3 hrs (95% confidence interval, 3.40-9.19) in the intermittent bolus group (p= .10). Less drug was administered in the intermittent group (mean, 0.02 ± 0.01mg/kg/hr) than by infusion (mean, 0.04 ± 0.01 mg/kg/hr;p< .001). Six patients (five in the infusion group and one in the intermittent group) developed persistent severe muscle weakness. In addition, six different patients (three from each group) had prolonged recovery >12 hrs.Conclusions:Our study suggests that recovery time after paralysis with continuous infusion is faster than that after intermittent bolus injection. Although more pancuronium was administered in the continuous-infusion group, recovery time was not prolonged as a consequence. It is uncertain whether pancuronium given by infusion increases the risk of persistent muscle weakness.