Although controversial, evidence from various experimental sources suggests a dysfunction/overactivity of the hemisphere dominant for speech - usually the left - in schizophrenia and of the other one in depression. Based on these lateralization hypotheses and on the dependency of regional cerebral blood flow on regional neuronal activity, the concept of pharmacopsychotherapy was proposed. The present double-blind pilot study tried to test this concept by cognitive training procedures tentatively relevant to the mechanisms of information processing of the left (analytical) and right (holistic) hemisphere, respectively. It was hypothesized that the outcome of patients putatively trained in mechanisms of the right hemisphere might be superior to that of those trained left in depression and the converse in schizophrenia. This hypothesis could not be verified. Apparent differences in the outcome of depressives were attributable to differences of independent variables (age, seriousness of illness).