The "single-point, single-dose" method for predicting individual maintenance dosage is examined. Data available in the literature on nortriptyline is used to illustrate the method. Optimal use of the method requires knowledge of the pharmacokinetics, especially the elimination rate constant, within the patient population requiring the drug. The method is applicable to intravenous and extravascular administration, when absorption is rapid relative to elimination, for those drugs whose disposition kinetics can be described by a linear one-compartment model. As a reasonable approximation, the optimal sampling time after the single test dose is 1.5 times the population half-life. Error in chemical analysis needs to be considered when applying the method. One method of evaluating the prediction and subsequent adjustment of dosage is discussed.