OBJECTIVE: To establish reference intervals for osteoarthritis (OA)-related biomarkers used in the Foundation for the National Institutes of Health (FNIH) OA Biomarkers Consortium Project. METHODS: A total of 129 'multijoint controls' were selected from 2722 African-American and Caucasian men and women in the Johnston County Osteoarthritis Project. The majority (79%) of those eligible (with biospecimens and baseline data) also had one or more follow-up evaluations 5-15 years later. Multijoint controls were selected to be free of radiographic hand, hip, knee and lumbar spine osteoarthritis (OA), to have no knee or hip symptoms, and minimal hand and spine symptoms at all available time points. Eighteen biomarkers were evaluated in serum (s) and/or urine (u) by ELISA. Reference intervals and partitioning by gender and race were performed with EP Evaluator software. RESULTS: Controls were 64% women, 33% African-Americans, mean age 59 years and mean body mass index 29 kg/m2. Three biomarkers were associated with age: sHyaluronan (positively), sN-terminal propeptide of collagen IIA (positively) and sCol2-3/4 C-terminal cleavage product of types I and II collagen (negatively). Exploratory analyses suggested that separate reference intervals may be warranted on the basis of gender for uC-terminal cross-linked telopeptide of type II collagen (uCTXII), sMatrix metalloproteinase-3, uNitrated type II collagen degradation fragment (uCol2-1 NO2) and sHyaluronan, and on the basis of race for uCTXII, sCartilage oligomeric matrix protein, sC-terminal cross-linked telopeptide of type I collagen and uCol2-1 NO2. CONCLUSIONS: To our knowledge, this represents the best and most stringent control group ever assayed for OA-related biomarkers. These well-phenotyped controls, representing a similar age demographic to that of the OA Initiative-FNIH main study sample, provide a context for interpretation of OA subject biomarker data. The freely available data set also provides a reference for future human studies.