Studies of the pharmacokinetics of vancomycin were conducted in a group of 28 patients with serious staphylococcal infection. Serum specimens were collected before and on 11 occasions after vancomycin administration. Serum concentration time data were fitted to a biexponential equation, using nonlinear regression analysis. A prolonged distribution phase with a half-life of 0.5 ± 0.3 h (standard deviation) and a central component volume of 9.0 ± 4.0 liters were demonstrated. Wide interpatient variation was observed in the terminal half-life which ranged from 3 to 13 h (mean, 6h) and in the distribution volume which ranged from 14 to 111 liters (mean, 39 liters). A correlation of 0.45 (Pearson product moment correlation coefficient) was found between vancomycin clearance and creatinine clearance. Multiple regression analyses demonstrated that 50% of the variance (R2) in the terminal half-life and vancomycin clearance could be explained on the basis of renal function, volume of distribution, age, weight, and sex. These observations suggest that adults with normal renal function should receive an initial dosage of 6.5 to 8 mg of vancomycin per kg intravenously over 1 h every 6 to 12 h. After 24 h, and through the period of therapy, trough and peak serum vancomycin concentrations should be monitored, and the dose and dosage interval should be changed to produce the desired peak (30 to 40 μg/ml) and trough (5 to 10 μg/ml) levels.