Summary: This paper views to empirically analyse the relationship of the direct effect and indirect effect which the attitude of high school students' volunteer activities might have on and to interpret the implications of the analysis results. For this, I schematized the formulaic model which explains the causal relationship of the relevant variables and set up the tree hypotheses for each of the direct and indirect effects. In the analysis, participation attitude which plays a role of an independent variable is composed of three subvariables of participation satisfaction, participation positiveness and participation autonomy. Self-exteem is the dependent variable. The parameter or intermediate variable is social support. For the test of hypotheses, I used the statistical methods of Anova, correlation analysis, and path analysis. The result is that all of the six hypotheses on direct and indirect effect were meaningful statistically. That is, the participation attitude of high school student's volunteer activities not only directly has an effect on their self-esteem but also has an indirect effect on their self-esteem through the parameter of social support.