This study employed a quasiexperimental control group design in a university setting to test the effect of a rater-training program on reducing social style bias in intragroup peer evaluations after controlling for ability based on GPA. Comparison of rating scores of the test group to the control group indicated minimal social style rating bias in the test group, whereas significant bias was exhibited in the control group. Implications for college instructors who use peer evaluations for grading in team projects are discussed. (Contains 2 figures and 3 tables.)