This study evaluated the efficacy of rational-emotive imagery as a component of rational-emotive therapy in reduction of college students' test anxiety. 11 volunteers met for 6 1-hr. group treatment sessions over a 3-wk. period. After 2 initial treatment sessions subjects were randomly assigned to groups given either rational-emotive therapy with rational-emotive imagery or rational-emotive therapy without imagery. Contrary to predictions, improvement between groups on self-report and performance measures was nonsignificant. Failure to obtain differences was attributed to similarities in content of treatment sessions and short treatment time. Combined groups reported significant improvement on all dependent measures. Although the study did not yield the predicted benefits of the imagery, results lend further support to the efficacy of rational-emotive therapy procedures in the reduction of test anxiety.