Numerous studies report about the relationship between the clinical effectiveness of amitriptyline (At) and the plasma level of this drug and of its most important metabolite, nortriptyline (Nt). These agents are therefore very frequently examined for clinical applications and for research programmes in specialized laboratories. The experience with antiepileptic drugs suggests the necessity of quality controls for antidepressants also. Therefore, five institutions in Western Europe performed two such experiments within a year. Two kinds of blood samples were sent for analysis: 1. plasma samples spiked with different quantities of At and Nt from an untreated subject; 2. plasma samples from patients treated with clinical doses of At. Each laboratory happened to use a different analytical method: TLC, HPLC, GC-NPD, GC-FID and GC-MS. The results clearly show the usefulness and the necessity of quality controls for this category of drugs as well. They form the basis for the improvement of the methods of each laboratory in particular. In this context, intralaboratory quality controls are also possible, if one disposes of two different methods. The findings of this study suggest that published reports on relationships between clinical and pharmacological parameters should be considered critically as to possible methodological bias.