OBJECTIVE: Time-trade-off (TTO) method is an important tool in health utility measurement but the contribution factors regarding its change especially on elder remains unclear. Purpose of this study was to explore the factors associated with TTO utility changes among old people. METHODS: A total of 288 male veterans aged above 65 years were longitudinally followed up one year for time-trade-off (TTO) utility. The data on WHOQOL-BREF (Taiwan version), socio-demographics, health status, health behaviors, home care, and health perception including rating scale (RS), self-rating health, and self-rating happiness were also collected. Wilcoxon rank sum test and Spearman correlation were used to select candidate of factor. Multiple regression analysis was performed to explore the contribution factors with the changes of TTO utility. RESULTS: The mean [SD] of TTO utility (0-1) for participants was 0.89 [0.25] at baseline and changed into 0.78 [0.34] after one year. Several factors at baseline were associated with the changes of TTO significantly (p <0.05) and independently including: age group, education level, rank when retired from army, stroke, respiratory diseases, number of office visits to a doctor, RS, self-rating health, physical domain and 9 facets of WHOQOL-BREF. After considerations of the collinearity of the variables and the simplicity of the models, variables entered in the final multiple regression models were: education level, stroke, RS, and "dependence on medical substances and medical aids" facet of WHOQOL-BREF (R2 = 0.12). CONCLUSIONS: Several variables were associated with TTO utility changes significantly and independently among old people at veteran home. Education level, stroke, RS, and ²dependence on medical substances and medical aids² facet of WHOQOL-BREF were associated factors of TTO utility changes in the final regression model. Due to the large variation of TTO utility changes the percentage of the variance accounted by the variables was small.