the purpose of this study was to compare the effectiveness of two treatments for alleviating math anxiety, rational emotive therapy and relaxation therapy. thirty-eight subjects were randomly assigned to three groups; rational emotive therapy, relaxation therapy and a no treatment control. at the end of three weeks, the data of the thirty subjects left were analysed. three different types of measures were taken to examine how the treatments affected the subjects. these measures were selfreport (mathematics anxiety rating scale and state-trait anxiety inventory), performance (answering taped math questions) and physiological (heart rate). the rational emotive therapy group was the only group to show an improved score on the self-report, mars test after treatment. the other measures showed no significant differences between treatment groups. the results were consistent with other studies showing ret to have the greatest effect on self-report measures of anxiety. the treatments were equal in clinical usefulness for treating math anxiety since there was no difference between any of the groups in performance.