In this study, we examined the relationship between time preference and depression tendency using two types of questionnaire data, which are Japan Household Panel Survey (JHPS) and Japanese Study of Aging and Retirement (JSTAR). For the empirical analysis, we applied the cross-lagged effect model and the synchronous effect model discussed in Finkel (1995), and we tested the Granger causality using the vector autoregressive model. The result showed that there was Granger causality in the direction of depression tendency to time discount rate in JHPS data. Moreover, in JHPS data, if the sample is divided by age, this causal relation was confirmed only with respect to over 50's. This result suggests that depression tendency influences time preference, which means that coping with the depression tendency makes it possible to change the time preference. Recently, approaches to depression tendency are progressing, and in particular, therapies called mindfulness are also adopted in general corporations and others. In other words, it can be interpreted that the analysis results showed that mindfulness leads to a change in time preference, that is, a decrease in time discount rate. With regard to such time preference changes, it can be said that it is desirable to use the viewpoint of virtue ethics in the framework of Bhatt et al. (2015).