Fifteen pairs of introductory statistics students, matched on initial test anxiety, were randomly assigned to eye movement desensitization (EMD) or control (finger tapping) conditions to test whether EMD effectively treats test anxiety and, if so, whether eye movement is the critical factor. Both groups had significant decreases in subjective units of disturbance during treatment, suggesting that another source of attenuation of elicited anxiety may be as effective as eye movement in reducing anxiety. There was a significant decrease in Test Anxiety Inventory Emotionality Scale scores from pretest to follow up for both groups, but mixed results on the Worry Scale and total anxiety scores.