This thesis consists of two studies. The first study was to investigate the relationships among family factors, educational backgrounds and assertiveness of college students. Attempts were also made to determine the variables which might effectively predict the assertiveness of college students The second study was to examine the effects of assertion-training. In the first study, 661 subjects were sampled by stratified judgmental method from the students population of National Taiwan Normal University. The dependent variables were the scores on the assertion, irrational beliefs, social anxiety and self-esteem. And the independent variables were the educational backgrounds (department, grade) and sex, family factors (father and mother´s attitude toward children, family climate, birth order, family location and the decision-maker of discipline). The data were analyzed by Pearson-product correlation, univariate two and three way analysis of variance, multiple regression, and Finn´s multivariate analysis of variance. The results indicated that educational backgrounds and sex, family location and mother´s attitude were significantly correlated with assertiveness of college students with low predict power. The scores of social anxiety, three of irrational beliefs (dependence, perfectionism, extreme superiority complex), social avoidance and distress together could effectively predict the assertiveness. In the second study, the pretest-midtest-posttest control group design was adopted. Twenty-four volunteered undergraduates who had low level assertiveness were randomly assigned to either treatment or control group. The treatment group received a weekly cognitive-behavioral approach group assertion-training for ten weeks, and the control group received no treatment. A final follow-up meeting was held two month after the last treatment session. The dependent variables were the scores of assertiveness, irrational beliefs, social anxiety, self-esteem and self-efficacy, and these scores were by Timm´s multivariate repeated measure analysis of variance, univariate analysis of covariance, Hotelling T^2and t-test. The results indicated that there were significant differences between the scores of treatment group and the control group except those at the pretest. It was, therefore, concluded that the treatment used in this study was effective in increasing assertive behavior and self-esteem and in decreasing irrational beliefs and social anxiety.