OBJECTIVES: To identify demographic factors that affect health preference elicited using the composite time trade-off (cTTO) technique in the China EQ-5D-5L Valuation Study. METHODS: In the China EQ-5D-5L Valuation Study, cTTO uses the traditional TTO for states better than dead and the lead time TTO for states worse than dead. A total of 86 EQ-5D-5L health states were grouped into 10 blocks with 10 health states in each group where one very mild state (i.e. only one level 2 is presented) and the worst state 55555 are included in every block. Demographics such as age, gender, employment status, health insurance, the attitude toward living is better than being dead (LBTD), and belief in afterlife were collected. Health preference was examined using mean, the number of states valued at 1 (non-trader), 0 (equal to being dead), -1 (potential censoring), the number of positive and negative scores, the difference between highest and lowest values, the very mild state had the highest value, and 55555 had the lowest value. Both univariate descriptive analyses and generalized linear models were used to explore the impact of demographics on health preferences. RESULTS: A total of CONCLUSIONS: We found that a few demographic indicators had significant impact on health preference. The impact needs to be adequately recognized in cost effectiveness analysis in China.