OBJECTIVES:  1)To determine the minimal important change (MID) of indirect utility score for patients with osteoarthritis (OA) over 3-year follow up; 2)Estimate the impact of comorbidity for patients reaching the MID. METHODS: 878 patients with symptomatic knee or/and hip OA of KHOALA cohort were included. Comorbidity were assessed by Functional Comorbidity Index (FCI). Utility score was assessed using SF-6D, and limitation in activities by WOMAC function score (0-100,0 is better), at baseline and 3-year. The MID of utility was assessed using standard error of measurement (SEM). According to MID, patients were classified as negative/positive change or unchanged. Two separate multinomial logistic models were fitted to determine predictors of positive or negative change over the MID threshold. Both models included sociodemographic characteristics and the mean difference of WOMAC score between baseline and 3-year. RESULTS: 650 OA patients completed the questionnaire at both times. The mean (SD) utility  was 0.664 (±0.110) at baseline and 0.667 (±0.110) at 3-year. Patients have on average 2.5 (±1.94) comorbidities and a signifcant (p<.0001) decrease of the WOMAC function score at 3-year. The MID of SF-6D utility score was 0.067: 147 patients classified with negative and 156 with a positive change. In the first model including the number of comorbidities, patients with a decrease of the WOMAC function score had an increased utility (OR=0.95; p<.0001). In the second model, patients with pneumologic (OR=1.88; p=0.03) or neurologic (OR=2.73;p=0.047) disease were likely to have improved utility, while patients having a psychiatric disease were less likely to have an improvement (OR=0.54;p=0.029). CONCLUSIONS: According to MID, about half of patients had a positive or negative change in their utility score. Compared to functional severity of OA, comorbidities have a relatively marginal impact on indirect utility score. This suggests that clinically, considering the functional severity of OA remains a first priority.