The lithium RBC/plasma ratio was determined in 59 patients during prophylactic lithium therapy -- 10 cycloid psychotics, 28 unipolars and 21 bipolars -- as well as in 20 patients during lithium therapy in phase of illness -- 8 cycloid psychotics, 5 unipolars and 7 bipolars. The lithium ratio was found to be normally distributed both among patients in phase of remission and those in phase of illness. No differences in lithium ratio were found between patients in phase of remission and phase of illness, respectively. During prophylactic lithium therapy, no differences were found in the lithium ratio between patients in different diagnostic subgroups. In a correlation matrix, the lithium ratio was found to correlate to lithium in RBC but not to age, sex, type of illness, weight, lithium dosage, duration of treatment or level of lithium in plasma. The lithium level in RBC was determined only by the lithium level in plasma. Thus, the lithium RBC/plasma ratio does not seem to be of diagnostic value in affective disorders.