Serum drug concentrations have commonly been described in terms of therapeutic ranges within which most patients have a therapeutic effect and a low incidence of toxicity. However, truly individualized drug dosage regimens cannot be developed without first setting a specific individualized target goal, such as a target serum drug concentration, at a desired target time after the dose (usually at a peak or trough), for each patient. For example, it is well known that the dosage of digoxin, or of any drug with a narrow therapeutic range, should somehow be individualized. One can begin this process by considering each patient as an individual, with his/her own individual need for the drug. If the need is small, so is the upper acceptable risk of toxicity. This would lead to a gently regimen, adjusted to the patient's body weight and renal function, to best achieve that specific target goal. Alternatively, if previous therapy has not sufficed and a significant or urgent need exists, then a higher goal may justifiably be selected, a greater risk of toxicity accepted, and a dosage regimen developed to meet that greater need. After such an individualized target goal is chosen, it should be achieved as precisely as possible. After the regimen is given, serum levels need to be measured and an individualized, patient-specific pharmacokinetic model should be made. Without the model, with only the raw serum level data, one cannot perceive the important exchanges that occur between serum and nonserum compartments of the drug, and we lack the precision given by the combination of the assay and the model to evaluate properly, optimally, the patient's clinical sensitivity to the drug. These concepts have been discussed here for digoxin, but they are general and apply to all drugs. This approach has also been applied to therapy with aminoglycoside antibiotics, vancomycin, lidocaine, theophylline, antiviral agents, a variety of anesthetic agents, psychiatric drugs, and anticancer agents.