HAVING EXPERIENCED THE PROBLEMS ASSOCIATED WITH (1) THE ABSENCE OF A SOUND KNOWLEDGE BASE, (2) THE LACK OF COST EFFECTIVENESS OF THE PROJECT-BY-PROJECT EVALUATION APPROACH, AND (3) THE INABILITY OF TRADITIONAL RESEARCH TO PROVIDE ANSWERS CONCERNING THE SYSTEM-WIDE COST EFFECTIVENESS OF SPECIFIC PROJECTS, ONONDAGA COUNTY DECIDED TO TRY A DIFFERENT APPROACH. THIS APPROACH IS BASED ON THE ASSUMPTION THAT EACH PROJECT IS PART OF A TOTAL CRIMINAL JUSTICE SYSTEM AND MUST BE EVALUATED IN TERMS OF THE TOTAL SYSTEM. IT ALSO ASSUMES THAT CLASSICAL EXPERIMENTAL DESIGNS MAY BE PREMATURE AT THE CURRENT LEVEL OF KNOWLEDGE ABOUT THE OPERATION OF THE CRIMINAL JUSTICE SYSTEM. INSTEAD, CONTRACTS ARE MONITORED BY ESTABLISHING BENCHMARKS AND MEASURING ACHIEVEMENTS AGAINST THESE BENCHMARKS. ORIGINALLY THE APPROACH WAS USED TO EVALUATE FIVE LEAA PROJECTS. SINCE MONEY AVAILABLE FOR EVALUATION WAS ONLY 3 PERCENT OF THE TOTAL AVAILABLE FOR THE PROJECTS, INHOUSE STAFF WAS TRAINED TO GATHERED THE EVALUATION DATA. EACH PERSON IN AN ADMINISTRATIVE CAPACITY WAS ON THE EVALUATION TEAM AND WAS RESPONSIBLE FOR THE COLLECTION OF A CERTAIN PORTION OF THE DATA. THE ESTABLISHMENT OF BASE RATES, THE TYPES OF DATA COLLECTED, AND THE ACTUAL OPERATION OF THE EVALUATION ARE DESCRIBED IN DETAIL. THE ULTIMATE PURPOSE WAS TO DETERMINE WHETHER THE PROJECT WAS DOING WHAT IT WAS SUPPOSED TO DO, AND (2) WHETHER WHAT IT WAS DOING WAS WORTHWHILE IN LIGHT OF MONEY AVAILABLE IN THE TOTAL SYSTEM. IT IS CONCLUDED THAT SUCH A SYSTEM-ORIENTED EVALUATION IS QUITE FEASIBLE FOR LOCAL GOVERNMENTS. APPENDED CHARTS DESCRIBE THE EVALUATIONS OF THE FIVE LEAA PROGRAMS AND ILLUSTRATE THE EVALUATION FRAMEWORK.