The metabolic ratio coefficient of debrisoquine (MR) was determined in l02 persons hospitalized with depression. No relations were found between the value of this coefficient and such patients’ characteristics as age, gender, diagnosis type (unipolar, bipolar, first episode depression), and the family history of affective disorders. Further analyses were done on a group of 77 patients, who were subsequently treated with antidepressive drugs that are known to be metabolised chiefly, or to a large extent, by CYP2D6. The undesirable side effects of treatment were not found to be significantly more frequent in the group of poor metabolisers (MR> 12.6), nor in the group of "increased risk" (intennediate metabolisers; 4 < MR < 12.6). A statistically insignificant but clearly visible tendency was found however, for the prescribed dosages to follow the metabolic profile of patients, i.e. patients with low MR values were given higher doses of drugs while the persons with higher MR values were given lower doses. The efficacy of antidepressive treatment was similar in all of the MR value brackets but the clinical evaluation revealed that the narrow group (6 persons) of poor metabolisers posed relatively more therapeutic problems than others did. The usefulness of the debrisoquine test in everyday clinical practice rests therefore in the possibility of defining this group, and modifying its treatment.