The selection of a starting dose for an antidepressant, and subsequent clinical titration to an appropriate therapeutic dosage, should be based on pharmacokinetic and pharmacodynamic principles. In the past decade, therapeutic monitoring of antidepressant drugs and use of pharmacokinetic principles have been shown to be an improvement over the dose-response approach. Endogenous (e.g., genetic metabolic phenotype, hepatic blood flow, and protein binding) and exogenous factors (e.g., smoking, dietary habits, concurrent medications) are capable of influencing physiological and pharmacokinetic variables in patients, accounting for the marked interindividual differences in the clearance rates of cyclic antidepressants. Interpatient variability for steady-state concentrations in plasma (Cpss) greater than 20-fold are observed at a fixed dose of imipramine (r2 = 0.525, df = 346, t = 19.541, P less than 0.0001) or doxepin (r2 = 0.506, df = 128, t = 11.403, P less than 0.0001). Analysis of doxepin in plasma vs estimated in oral clearance for 61 patients demonstrates a significant decline in oral clearance as a function of Cpss. At doses approaching the upper range recommended for the treatment of depression, Cpss appear to approach, in at least a few individuals, the maximum metabolic capacity of the patient (Vmax), leading to greater-than-expected increases in concentrations for a given dosage increment. Significant alterations in oral clearance are observed when medications are administered concomitantly. A greater-than-threefold difference in mean oral doxepin clearance rates is observed between two groups of patients receiving additional medications that are either inducers or inhibitors (P less than 0.0001, df = 32, t = 6.687). Pharmacokinetic principles defining and explaining the determinants of oral clearance can provide the clinician with a greater insight into the reasons for therapeutic failure and toxicity.