In 1995 all systemic lupus erythematosus (SLE) patients in the county of Funen were retrieved from four separate and independent sources as part of an 8-year prospective study to determine the pattern of disease activity and damage accumulation in a community based lupus cohort of predominantly Scandinavian ancestry. Incident cases were subsequently identified by surveillance of these sources. Established and new cases underwent annual, structured interviews, clinical examination and blood sampling. The Systemic Lupus Erythematosus Diseases Activity Index SLEDAI and Systemic Lupus International Collaborating Clinics SLICC scores were calculated. Flares were defined as modified - SLEDAI >or= 4. The annual flare rate in definite SLE (D-SLE) was 0.21 (95%CI 0.18-0.24) versus 0.03 (95%CI 0.01-0.07) in incomplete SLE (I-SLE). Forty-three per cent of the entire study population had no disease exacerbations. Infections requiring hospital admission and thrombocytopenia were significantly more frequent among patients with relapsing disease (p < 0.04-0.01). Patients with flares had slightly shorter disease duration and were younger at disease onset than patients with a quiescent course. The most recently diagnosed patients had the lowest annual rate of damage accrual. According to flare rate, two major subsets of almost equal size were identified - one having a long quiescent course, the other exhibiting relapses alternating with remissions. An increased risk of flares was associated with short disease duration and younger age at disease onset, infections requiring hospital admission and thrombocytopenia. Temporal damage increment was the lowest in the most recently diagnosed patients.