The pharmacokinetics of neostigmine in patients with normal renal function (n = 8) were determined and compared with those of patients undergoing renal transplantation (n = 6) or bilateral nephrectomy (n = 4). All patients were anesthetized with nitrous oxide and halothane. d-Tubocurarine was infused at a rate sufficient to maintain 90 per cent depression of twitch tension. Ten to 15 minutes prior to the end of operation and anesthesia, the d-tubocurarine infusion was terminated and neostigmine, 0.07 mg/kg, and atropine, 0.03 mg/kg, were given by infusion over a 2-min period. Concentrations of neostigmine in blood drawn periodically during the following four hours were determined by gas-liquid chromatography and the data fitted to a two-compartment pharmacokinetic model. In anephric patients elimination half-life (181 ± 54 min, mean ± SD) was significantly prolonged when compared with comparable values for patients with normal renal function (80 ± 48 min). Total serum clearance was significantly decreased from 16.7 ± 5.4 ml/kg/min in patients with normal renal function to 7.8 ± 2.6 ml/kg/min in anephric patients. Neostigmine pharmacokinetics following renal transplantation were not different from those in patients with normal renal function. It is concluded that renal excretion accounts for 50 per cent of neostigmine clearance and, in the absence of renal function, the serum half-life of neostigmine is prolonged, similar to that of d-tubocurarine.