A DECISION MODEL IS A SET OF WEIGHTED VARIABLES OR ELEMENTS OF INFORMATION THAT, IF PRESENT IN A CRIME REPORT, WILL ENABLE THE CASE OUTCOME TO BE ACCURATELY PREDICTED. THE STANFORD RESEARCH INSTITUTE FIRST EXPLORED THE USE OF SUCH CHECKLISTS AND FOUND THEM TO BE RELIABLE FOR BURGLARY AND ROBBERY. THIS MINNESOTA STUDY TESTED THE MODELS IN FOUR LAW ENFORCEMENT AGENCIES SERVING MEDIUM-SIZED CITIES OF 20,000 TO 60,000 PEOPLE. THE FOUR CITIES WERE MAPLEWOOD AND RICHFIELD (METROPOLITAN AREA) AND ROCHESTER AND ST. CLOUD (OUTSTATE). DATA COLLECTION COVERED A 3-MONTH PERIOD. ALL CLEARED AND UNCLEARED CASES FOR ROBBERY WERE REVIEWED, USUALLY COVERING THE YEARS 1974 THROUGH 1976. FOR BURGLARY, RESEARCHERS EXAMINED ALL CLEARED CASES AND A SAMPLE OF UNCLEARED CASES. ALSO REVIEWED WERE 270 ROBBERY CASES; THE ROBBERY DECISION MODEL APPLIED TO 210 OF THEM. THE DEGREE OF ACCURACY ACHIEVED WAS 87 PERCENT. FOR BURGLARY 401 CASES WERE USED AND THE DEGREE OF ACCURACY WAS 91 PERCENT. THUS THE DECISION MODELS FOR BOTH ROBBERY AND BURGLARY PROVED SUCCESSFUL AND USEFUL FOR PREDICTIONS, AND A KEY RECOMMENDATION WAS THAT ROBBERY AND BURGLARY DECISION MODELS BE CONSIDERED FOR IMPLEMENTATION BY LAW ENFORCEMENT AGENCIES IN MINNESOTA. TABULAR DATA ARE INCLUDED. (MJW)