Although the manufacturer of the polyclonal fluorescence polarization immunoassay (FPIA) for tricyclic antidepressants (TCA) only recommends its use in the diagnosis of overdose, the assay is nevertheless widely used in therapeutic drug monitoring. Using plasma samples from 337 patients taking one of eight different tricyclic antidepressants, the authors investigated the performance of the TDx assay procedure for eight different TCAs by comparison to specific high-performance liquid chromatography (HPLC) assay methods. The regression correlation between the TDx assay value and that for active tricyclic measured by HPLC was poor (r2< 0.9) for amitriptyline, clomipramine, dothiepin, and doxepin. The regression line for amitriptyline also had a significant positive y-axis intercept. Moreover, the TDx method overestimated the concentration of active drug to an extent that varied considerably between different TCAs and within the usual therapeutic range for a single TCA. The authors conclude that the TDx assay is probably satisfactory for routine TDM of desipramine, imipramine, nortriptyline, and trimipramine. However, it significantly overestimates therapeutic concentrations of amitriptyline, clomipramine, dothiepin, and doxepin. The use of TDx and HPLC assay methods by different laboratories for sequential therapeutic drug monitoring of TCAs in the same patient may confuse physicians and confound dose adjustment and patient management. Although their study shows that the TDx assay can give satisfactory therapeutic drug monitoring results for some drugs, the authors conclude that its use should be restricted to the evaluation of overdose as recommended by the manufacturer.