Twenty test-anxious subjects were randomly assigned to four systematic desensitization conditions, representing two levels each of therapist warmth and status. Ratings by treated subjects and independent judges confirmed that the experimental warmth and status manipulations had been successful. Comparisons with an untreated, high test anxiety control group (N = 10) indicated that the treated subjects evidenced significantly greater reductions in both test and trait anxiety. Within the treated subjects, changes in both test and trait anxiety were greatest in the two higher warmth conditions. There were no statistically significant effects of therapist status. Subjects' ratings of satisfaction with treatment and likelihood of return to desensitization treatment should new problems occur, were also highest in the two higher warmth conditions. Reasons for the differential effectiveness of the warmth and status factors are discussed, and future research directions in the areas of behavior therapist personality and role variables are suggested.