This study was conducted to investigate whether cognitive variables (as measured by six self-report inventories) covaried with changes in behavior following group assertiveness training as postulated by several assertiveness training theorists and practitioners. Statistical analysis of the data did not support any of the hypotheses. No significant differences were found between the groups on any of the self-report measures or the behavioral ratings, although experimental group subjects consistently scored higher (in the predicted direction) on all self-report measures or were rated higher than placebo group subjects. The hypotheses that experimental group subjects would exhibit significantly higher levels of self-reported assertiveness, self-acceptance, self-esteem, and self-confidence than placebo group subjects were rejected. The hypothesis that the experimental group would exhibit a significantly greater shift toward self-reported internal locus of control than the placebo group was rejected. The hypothesis that the experimental group would exhibit significantly larger decreases in levels of self-reported anxiety than the placebo group was also rejected. Furthermore, no support was found for the hypothesis that experimental group subjects would exhibit significantly higher levels of objectively-rated assertive behavior than placebo group subjects. An examination of the data revealed substantial improvements on each of the six self-report measures for subjects in both groups. These data were treated statistically with a one-tailed t-test for correlated means. The t-test data indicated that both groups improved significantly on all cognitive variable measures from pretest to posttest.