Serum hyaluronan (HA) concentration was quantified using an ELISA-based assay with biotinylated hyaluronan binding proteins, and correlated with the clinical and laboratory variables in 100 consecutive rheumatoid arthritis (RA) patients (mean +/- SD age and duration of disease of 50.1 +/- 12.5 and 7.9 +/- 6.6 years respectively). Thirty-four patients received prednisonole at an average dose of 5.0 mg/day. The correlations were good between the serum HA level and the joint swollen scores (r = 0.26, p = 0.04), joint space narrowing scores (r = 0.25, p = 0.03), joint erosion scores (r = 0.26, p = 0.03), and erythrocyte sedimentation rate (r = 0.31, p < 0.01) in RA patients who did not take prednisolone. These correlations were diminished in those who received prednisolone, although their disease was more severe. It might be possible that corticosteroids could decrease inflammation of the joint, thus interfering with the correlations. It was concluded that the serum HA level is a useful marker for the activity and severity of disease in patients with RA.