Steady-state plasma levels of nortriptyline and E- and Z-10-OH-nortriptyline were determined in 55 depressed patients during long-term treatment. Dose-corrected steady-state levels varied by a factor of 20 for nortriptyline, a factor of 7 for E-10-OH-nortriptyline (sum of enantiomers), and a factor of 12 for Z-10-OH-nortriptyline (sum of enantiomers). The E-10-OH-nortriptyline levels were higher than the corresponding nortriptyline levels in about 50% of the patients and the nortriptyline/E-10-OH-nortriptyline ratio ranged from 0.27 to 4.8. In contrast to E-10-OH-nortriptyline, the steady-state levels of Z-10-OH-nortriptyline correlated significantly with the nortriptyline levels (rs = 0.68, n = 55, p less than 0.001) and the nortriptyline/Z-10-OH-nortriptyline ranged from 1.7 to 10. Patients on concurrent treatment with perphenazine or benzodiazepines had higher nortriptyline and nortriptyline/E-10-OH-nortriptyline ratios than patients taking lithium or no other psychotropic drugs. A sparteine test was carried out in 22 patients and the sparteine metabolic ratio correlated significantly with the dose-corrected steady-state levels of nortriptyline (rs = 0.62, p less than 0.01) and E-10-OH-nortriptyline (rs = -0.52, p less than 0.02) and particularly well with the ratio nortriptyline/E-10-OH-nortriptyline (rs = 0.83). The genetic variability in the sparteine/debrisoquine P-450 isozyme appeared to be clearly more important for the interindividual variation in 10-hydroxylation of nortriptyline than the possible interactions with concurrent medication.