A competitive ELISA with an inhibition step for the determination of serum keratan sulphate (KS), was validated using bovine corneal KS as standard. The method was shown to be in statistical control. The recovery was 85% (S.D. = 2.2%), hence the obtained results had to be corrected. The limit of detection (LOD) was 1.4 μg m l^-1. Intra-assay CV was 3.0% (n = 15) for samples used for method evaluation and 6.0% (n = 18) for serum samples. Inter-assay CV was 7.0% for high KS level samples (C (KS) = 31.2 μg m l^-1) and 16.0% for low KS level samples (C(KS) =5.9 μg m l^-1) based on the S.D.(y/x) from the method evaluation function and 10% (high KS level) and 16% (low KS level) based on serum samples. Validation of the method was performed using the serum samples from 102 healthy individuals. There was a statistical difference in serum KS levels between men and women, but no age related differences. Tentative biological reference intervals for men and women were established. It was also shown that due to the large between-subject variability in KS levels, the preferable monitoring strategy is to use individuals as their own controls where possible. (C) 2000 Elsevier Science B.V.