The evaluation used the proposed guideline scoring of various characteristics of the inmate to determine a key factor score. This score represented a category in a matrix designed to incicate an appropriate length of time to be served for a given offense. By calculating key factor scores, the probable guideline release dates of a sample of about 400 Wisconsin inmates convicted of armed robbery and burglary charges in 1974 and 1975 were estimated. The guideline release dates calculated by scoring this sample of inmates were compared to the real release dates of the subject. The proposed guidelines did not model past parole release decisions for the offenses of armed robbery and burglary largely because of the extreme range of scoures in each guideline category. Additional variables shown to have significant effects on release decisions but not included in the guideline matrix were also responsible for the lack of agreement between guideline and real release dates. A series of guideline revisions is presented, with the first priority revision involving the refinement of an offense seriousness dimension. The guidelines in the final revision also regularize length of time served with offense seriousness. The final revised guidelines were found to be equitable and workable for the two offenses, provided that they are implemented in a standardized and reliable manner. The data collection instrument and a discussion of the effect of personal variables are appended. Tabular data and a bibliography (ca. 50 references) are provided.