Despite its long use in clinical medicine, protamine concentrations and pharmacokinetics in humans have not been reported. The occasional reoccurrence of anticoagulation after protamine reversal of heparin led us to hypothesize that protamine plasma concentrations decrease rapidly. We developed a method for the measurement of protamine in plasma. Eighteen fit volunteers gave their consent to receive 0.5 mg/kg protamine sulfate administered IV by an infusion pump over 10 min. Heart rate, mean arterial blood pressure, and cardiac output, all measured noninvasively, were recorded and blood samples obtained during and after protamine infusion. Blood plasma was subjected to solid-phase extraction and high-performance liquid chromatography. The administration of protamine was associated with no significant changes in heart rate, mean arterial blood pressure, or cardiac output. Plasma protamine concentrations decreased rapidly, becoming nondetectable within approximately 20 min. Protamine elimination differed significantly between men and women: men had significantly larger areas under the concentration versus time curve. Model-independent pharmacokinetic analysis revealed median (range) values as follows: volume of distribution at steady state, 12.3 (6.9-63.1) L; clearance, 2.2 (1.1-12.1) L/min; and t1/2, 7.4 (5.9-9.3) min. Concentration versus time plots revealed an atypical pattern inconsistent with usual exponential models. The Schwartz-Bayesian criterion identified a one-compartment Michaelis-Menten model and a two-compartment exponential model with irreversible binding as performing better than conventional one- or two-compartmental exponential models; however, performance errors were large with both Michaelis-Menten and exponential models. All models described rapid decreases in protamine blood concentrations.