There is a need for a simple depression questionnaire also capable of assessing the severity of depression. The Depression Scale (DEPS), has been a very popular self-rating depression questionnaire in Finland for >15 years. Our aim was to examine whether the DEPS has the ability to differentiate clearly defined levels of depression in primary care patients. Primary care patients aged 18-64 years completed a postal questionnaire including the DEPS. All screen-positive subjects and every 10th screen-negative subject were invited for interview using the Present State Examination (PSE) as the gold standard. Complete DEPS score was available for 410 patients. Descriptive statistics of the DEPS in the six diagnostic PSE classes were computed. Four of the PSE classes were selected for further analyses of depression severity. Receiver Operating Characteristic curves, sensitivity, specificity, ideal cut-off points and area under the curve were calculated. The ability of the DEPS to differentiate levels of functioning was also evaluated. The DEPS identified three groups of patients: those with no psychiatric symptoms, those with some depressive symptoms and those with clinical depression. The margins between the levels were thin: the ideal cut-off point for clinical depression was 11/12 and for any level of depression 9/10. The DEPS was also able to differentiate three levels of functioning. The DEPS has some ability to identify severity of depression in primary care patients. Further research with larger unscreened material is called for.