Over the years of continuous problem with performance rating systems has been the leniency and the non-comparability of marks assigned by different evaluators. By utilizing a computer, a method has been developed which overcomes this problem. In this new method the evaluators must compare their ratees to other specific ratees (anchors) who are under other evaluators. All ratees receive their scale value based on their relative position to the anchoring points that were used by the evaluator. A trial of the method was made on ten groups, each composed of 12 graduate students. Each group has two evaluators. The characteristics rate were Industry, Academic, Ability, Judgment, and Cooperation. These were considered as relevant characteristics to the "job" of being a student. The groups had been relatively intact for approximately one year prior to the evaluations. A comparison of the results of using the standard fitness report rating method (RAW) and the man-to-man method (MM &MMQ) revealed that the man-to-man method was superior based on certain statistical qualities. The man-to-man method resulted in a greater spread of scores and, more importantly, resulted in higher inter-rater agreement than the standard rating method. An outside criterion of Quality Point Average was available for the "Academic Ability" scale. Both the standard methodology and the man-to-man methodology produced rating values which were highly related to this outside criterion, .68 and .71 respectively.