Self-assessment scales have long been used in psychiatric research even if their validity has often been questioned, one reason being poor the concordance of expert ratings. In clinical practice the use of rating scales is restricted, since they are considered to be time-consuming and perhaps even to disrupt the clinician's rapport with the patient. In the present study, a self-assessment scale, the CPRS Self-rating Scale for Affective Syndromes (CPRS-S-A), was constructed by re-phrasing in a self-rating format 19 items from the original Comprehensive Psychopathological Rating Scale (CPRS) and covering depression, anxiety and obsessional symptoms. In a group of 30 patients with depression syndromes and anxiety syndromes, the CPRS-S-A and the original CPRS were both used on 2 occasions. The patient's Global Assessment of Functioning scores ranged from 30 to 76 (mean 58), which suggests a moderate severity of illness, as does the fact that the majority were outpatients. There was a high degree of concordance between the instruments for most items and for the scores on the subscales for both diagnostic groups (i.e., the Montgomery-Asberg Depression Rating Scale and the Brief Anxiety Scale, which are both subscales drawn from the CPRS). The time taken to complete the CPRS-S-A varied from 5 to 30 min (mean 19 min for depressive and 16 min for anxiety patients on the first occasion, 13 min for both groups on the second), and the self-rating procedure was readily accepted by both groups of patients. The CPRS-S-A would thus seem to be a promising instrument for quantitative rating of symptoms in ambulatory patients, both in clinical practice and in research.