Thiopental was used in long-term infusion (3–4.5 mg · kg−1· h−1during 4–8 days) to protect the brain from injury following trauma. Thiopental plasma concentrations were measured during infusion (48 patients) and after infusion (14 patients) to determine the kinetics of the drug in continuous infusion. All mean values were mean ± SD. Steady state concentrations (C18) were 31.8 ± 10.7 mg/1 for an infusion rate of 3.05 ± 0.37 mg · kg−1· h−1and 48.9 ± 14.6 mg/1 for a rate of 4.2 ± 0.3 mg · kg−1· h−1. Corresponding steady state clearance decreased when C18increased, indicating possible saturation of the metabolic enzymatic system. Michaelis-Menten kinetics were confirmed by postinfusion data that give, for higher C18, a nonlinear decay of log Cversustime. First-order kinetics were only obtained with C18, below 30 mg/1. The maximum rate of elimination (Vm) was 1.76 ± 1.15 mg · I−1· h−1(n = 11), and the Michaelis constant (Km) was 26.7 ± 22.9 mg/1 (n = 11). Hepatic enzyme saturation was between 35 and 85%. The volume of distribution at steady state was 4.35 ± 1.83 1/kg (n = 11). Apparent half-lives of elimination were between 18 and 36 h at the end of infusion, and predicted terminal half-lives were 10.15 ± 5.43 h (n = 11). Phases of burst-suppression were observed on electroencephalographic traces for concentrations greater than 40 mg/1. The authors' results suggest that a continuous infusion at a dose of 4 mg · kg−1· h−1induces EEG changes consistent with a near-maximum reduction in cerebral metabolism. Because of the thiopental Michaelis-Menten kinetics at doses above 4 mg · kg−1· h−1, the authors suggest that thiopental plasma concentrations be measured and/or the drug effect be measured with the EEG to prevent excessive thiopental overdosage, causing a prolonged recovery time.