Main and alternative diagnoses were applied to five hundred consecutive admissions. A group of patients was isolated with an affective disorder as the main diagnosis and

schizophrenia as the alternative diagnosis ('A-S Mixed'). In symptoms and outcome this group was intermediate between groups of patients where there was no alternative diagnosis, or where the main and alternative diagnoses were both of an affective disorder or of schizophrenia. Further analysis showed that the use of alternative diagnoses makes it possible to identify those diagnoses which are creating uncertainty and confusion between categories.