BACKGROUNDS:The EuroQol Group Five-Dimensional (EQ-5D) instruments have been validated in China for measuring health-related quality of life (HRQoL) and are increasingly being used in health economic studies. However, there is paucity in the literature documenting long-term changes in the EQ-5D results in the Chinese populations. This study aims to identify such changes and their determinants using the EQ-5D-3 L instrument. METHODS:Data were obtained from the National Health Services Surveys in China, which included the EQ-5D-3 L since 2008. We compared the differences between the 2008 and 2013 surveys in the percentage of reported problems, visual analogue scale (VAS) scores, and the EQ-5D-3 L utility index derived from the national value sets. Factors associated with population changes in these EQ-5D results were identified using logistic, linear and Tobit regression models, respectively. RESULTS:Compared with 2008, reported problems in self-care (3.3% vs 3.1%), usual activities (4.8% vs 4.6%) and anxiety/depression (6.4% vs 5.3%) decreased, whereas reported problems in mobility (5.1% vs 5.9%) and pain/discomfort (9.3% vs 12.6%) increased significantly (p < 0.05) in 2013. The regression models revealed a rise (β = 1.61, p < 0.001) in VAS scores, but a slight drop (β = - 0.01, p < 0.001) in utility index in 2013 compared with 2008 after controlling for variations in demographic, behavioral, socioeconomic and residential variables. But the effect sizes of the changes over time (estimated by "average change divided by baseline standard deviation") did not reach the threshold of clinical importance after adjustment for variations in other factors. Higher socioeconomic status (in terms of education, income and residential location) was associated with better EQ-5D-3 L results. CONCLUSION:The changing trend (decrease) of the utility index is contradictory to that (increase) of the VAS scores, although neither is deemed clinically important. It is evident that socioeconomic and regional disparities in HRQoL exist in China.