A population based cohort of patients with systemic lupus erythematosus (SLE) was recruited from a for epidemiological purposes representative Danish region. Patients were ascertained from 4 different sources with a high degree of completeness as estimated by using capture-recapture analysis. The diagnosis was verified by means of case records, patient interviews, and clinical examinations. Patients were classified according to the 1982 revised ACR criteria as Definite SLE (D-SLE) fulfilling > or = 4 criteria and Incomplete SLE (I-SLE) with < 4 criteria. As of January 1, 1995, the point prevalences of D-SLE and I-SLE were 21.7 and 5.2 per 100000 respectively. The cohort comprised 98% white Europeans. The annual incidence of D-SLE increased from 1.0/100,000 to 3.6/100,000 during the study period 1980-94. D-SLE patients had clinical profiles comparable to other Western European study populations. The I-SLE subclass had milder disease manifested by lower criterial load and absence of cerebral and kidney involvement. Tissue and organ damage expressed as SLICC-score increased in the order of live I-SLE, live D-SLE, and deceased patients.