Information on steady-state concentrations of parent tricyclic antidepressants (TCAs) and their major metabolites in plasma is useful in ascertaining compliance, for possible pharmacokinetic changes during longer treatment, and for prospective individualized dosing procedures. Adequate response can be maintained when there are relatively small fluctuations in drug concentration from visit to visit, but large fluctuations increase the liklihood of relapses during the acute-treatment phase and recurrences afterwards. Changes in the individualized in dosage regimens complicate measurement of concentrations in plasma at steady state. Longitudinal examinations of concentration/dosage (L/D) values in our three-year imipramine (IMI) maintenance study reveal that patients having large intra-individual fluctuations can be classified as noncompliers, likely to have recurrences. The time course and magnitude of L/D values reflect higher accumulation of both components, IMI and desipramine (DMI), in plasma. Evidently, dose-dependent kinetics lead to higher steady-state concentrations in plasma than previously observed with similar dosages in shorter-term treatment. With amitriptyline (AMI), in a 12-month study, the mean total concentrations of both AMI and nortriptyline (NT) progressively increased, by 22% and 33%, respectively. The pharmacokinetic linearity of AMI and (or) NT, i.e., (L/D)high/(L/D)low, is maintained over much wider dosage and age range than with IMI or DMI. We advocate concurrent use of the L/D method and co-administration of riboflavin for identification of noncompliers. We describe our current experience with various analytical procedures in this regard, concluding that high-performance liquid-chromatographic methods, with appropriate selection among ultraviolet, enhanced fluorescence, and electrochemical detectors for each TCA under specified therapeutic conditions, are most suitable and versatile, having superior analytical parameters. A suitable alternative procedure is gas-chromatography (N/P mode). Although immunoassays (EMIT, radiochemical) are most convenient for toxicological screens, their significant cross reactivities with several phenothiazines and the detection limits, requiring higher concentrations for EMIT, restrict their usage during research and (or) clinical monitoring.