The continuing morbidity of patients with vasculitis, despite the improved prognosis with aggressive therapy, underlines the need for accurate disease assessment. We have devised a clinical index of disease activity, and evaluated its use in several forms of necrotizing vasculitis. The weighted score is based on symptoms and signs in nine separate organ systems. Disease features are only scored if they are attributable to active vasculitis. The Birmingham Vasculitis Activity Score (BVAS) was compared with two other published vasculitis activity scores, with the physician's global assessment (PGA), with outcome, and with serological markers of disease activity. In a cross-sectional study of 213 consecutive patients with different forms of vasculitis, all 107 vasculitis patients who were judged completely well on clinical assessment had a BVAS score of 0. Twenty-two patients with active vasculitis prior to treatment had a median score of 7.5 (range 4-30) and 69 with active disease on treatment had a median score of 10 (1-29). Of the 12 who died, median score immediately prior to death was 20.5 (9-30). In a serial prospective study, 30 cases had documented episodes of active disease. During periods of disease activity, the median BVAS values were significantly higher than in remission (15 [range 3-32] vs. 0 [0-2], p < 0.001); the same was true for CRP values (80 [9-361] vs. 13.5 [5-68], p < 0.001). This was not true for erythrocyte sedimentation rate (ESR), haemoglobin (Hb) or von Willebrand factor (VWF).(ABSTRACT TRUNCATED AT 250 WORDS)