BACKGROUND Sclerostin alters bone formation. The precise and reproducible measurement of sclerostin concentrations in biological samples is important for assessment of metabolic bone disease. We determined sclerostin concentrations in serum and plasma using two commercially available ELISA. METHODS We measured sclerostin concentrations in serum or heparin-plasma obtained from 25 normal human subjects using two commercial ELISA available from Biomedica Medizinprodukte GmbH and TECOmedical AG. RESULTS With the Biomedica assay, serum sclerostin concentrations were 0.99 ± 0.12 ng/ml (mean ± sem), and plasma concentrations were 1.47 ± 0.13 ng/ml (paired t test, P < 0.001). With the TECO assay, serum sclerostin levels were 0.71 ± 0.05 ng/ml, and plasma sclerostin concentrations were 0.80 ± 0.06 ng/ml (paired t test, P < 0.001). Serum and plasma sclerostin concentrations were significantly different when determined by the two assays (serum, P = 0.015; plasma, P < 0.001). Recovery of added recombinant sclerostin to serum was less than expected with both Biomedica and TECO assays (P < 0.001, paired t test). CONCLUSIONS The concentrations of sclerostin in serum and plasma are different when determined by the two assays. Serum or plasma sclerostin concentrations with current assays should be interpreted with caution. The data suggest that the same assay should be used for comparing groups of patients or patients being followed longitudinally. Standardization of sclerostin assays is required before being introduced into general clinical laboratory use.