OBJECTIVES: To map the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) onto the EuroQol 5-Dimension (EQ-5D) utility index in patients with knee or hip osteoarthritis in a clinical trial (N=609); the United Kingdom algorithm was used to calculate EQ-5D utility scores. METHODS: WOMAC was evaluated as a continuous variable using total scores from baseline and week 12 by summarizing all items and then converting to a 0-10 scale. Repeated measures models (RMM) that considered all available data across time were implemented by imposing a linear relationship with EQ-5D score as the outcome and WOMAC total score as the predictor. A model with WOMAC total score as a categorical variable was evaluated in a sensitivity analysis. In secondary analyses, Ordinary Least-Square regression (OLS) with WOMAC averaged over time as predictor was performed, and a RMM model with an additional quadratic term was evaluated. Cross-validation was used to compare RMMs. RESULTS: In the main model, WOMAC total score as a continuous variable showed sizable and significant correlation with EQ-5D (-0.58; P<0.0001) with a relationship of EQ-5D = 0.9202 − 0.07055×WomacTotalScore=-0.34) produced practically the same results as RMM. While categorical analysis suggested a linear approximation is appropriate when allowing for natural sampling variation, it also suggested that introducing a quadratic term may be justified. Cross-validation indicated that stability and robustness of these two models were not dependent on the number of subsamples. The mean squared difference between observed and predicted values was slightly lower (better) with the quadratic model than with the linear model (0.057 vs. 0.059).  CONCLUSIONS: Mapping WOMAC total scores to the EQ-5D utility scores showed reasonable fit. While the linear function provided a more parsimonious adoption, further evaluation may be needed to determine whether linear or quadratic models should be applied to particular cases.