IN ORIGINAL TESTS OF DETERRENCE THEORY, IT WAS FOUND THAT THE PROBABILITY OF APPREHENSION WAS NEGATIVELY RELATED TO THE CRIME RATE, BUT ONLY WHEN THE DEGREE OF CERTAINTY HAD REACHED A MINIMAL LEVEL. THE LATER STUDY ATTEMPTED TO ESTABLISH THE DIRECTION OF CAUSALITY USING INFORMATION FOR TWO POINTS IN TIME. DATA SHOWING THE ARREST CLEARANCE AND CRIME RATES FOR EACH MUNICIPALITY AND COUNTY IN FLORIDA WERE USED. THE PROBABILITY OF APPREHENSION WAS UTILIZED AS ONE VARIABLE BECAUSE IT PREDETERMINES THE MAXIMUM PROBABILITY OF ANY LEGAL SANCTION. CRIME VOLUME DATA FROM THE FEDERAL BUREAU OF INVESTIGATION WERE EMPLOYED BECAUSE OF THE RELATIVE RELIABILITY. TWO ANALYSES WERE CONDUCTED, ONE TO SEE IF AN INCREASE IN THE SIMPLE PROBABILITY OF APPREHENSION CAUSES A DECREASE IN CRIME, THE OTHER TO SEE IF AN INCREASE IN POLICY EFFICIENCY WOULD LEAD TO A DECREASE IN CRIME. THE RESULTS SHOWED THAT A DETERRENT EFFECT OCCURS ONCE THE CERTAINTY OF ARREST REACHES A CRITICAL LEVEL. HOWEVER, THE SAMPLE USED IN THE COMPARISON MAY BE TOO SMALL AND MAY HAVE CAUSED A DISTORTION IN THE RESULTS. STATEMENTS ABOUT DETERRENT EFFECTS, ESPECIALLY THOSE BASED ON CORRELATIONAL STUDIES USING OFFICIAL POLICE DATA, MUST BE ADVANCED WITH CAUTION AND IN VERY SPECIFIC TERMS. CHARTS AND REFERENCES ARE PROVIDED. (BAC)