Serum concentrations of antigenic keratan sulphate determined by an enzyme linked immunosorbent assay (ELISA) with a monoclonal antibody were studied in patients with rheumatoid arthritis (RA), osteoarthritis, ankylosing spondylitis, other inflammatory diseases, and a large control group of women without arthritis. Mean keratan sulphate concentrations were low in 117 women with RA compared with 227 female control subjects matched for age drawn from a community survey. There were significant correlations between serum keratan sulphate concentrations in patients with RA and serum C reactive protein and the erythrocyte sedimentation rate. Serum keratan sulphate concentrations were also low in 29 men and women with ankylosing spondylitis and 29 patients with arthritis and high concentrations of C reactive protein. In 98 women undergoing an operation for benign breast disease there were decreases in serum keratan sulphate concentrations after the operation which correlated with doses in serum C reactive protein. No differences were found in keratan sulphate concentrations in 137 women with osteoarthritis compared with controls. Within the group with osteoarthritis there were no differences for the various joint groups and there was no obvious correlation with radiographic severity or progression. These findings suggest serum keratan sulphate is unlikely to be useful as a diagnostic marker in osteoarthritis or RA but indicate a role for inflammation in the regulation of cartilage loss.