A QUESTIONNAIRE WAS CONSTRUCTED TO MEASURE RISK PERCEPTION. USING OFFICIAL STATISTICS AS A BASELINE, DELINQUENTS AND NONDELINQUENTS ARE FIRST COMPARED AS TO HOW THEY PERCEIVE ARREST AND CONVICTION RATES FOR VARIOUS CRIMES. SECOND, THEIR PERCEPTIONS OF THE RISK TO THEMSELVES OF ARREST AND CONVICTION FOR THREE HYPOTHETICAL OFFENSES ARE COMPARED. THE FOLLOWING THREE MAJOR HYPOTHESES ARE TESTED: DELINQUENTS PERCEIVE LAW ENFORCEMENT TO BE LESS EFFECTIVE THAN DO NONDELINQUENTS; DELINQUENTS SEE THEMSELVES AS MORE LIKELY TO VIOLATE THE LAW IN HYPOTHETICAL SITUATIONS THAN DO NONDELINQUENTS; AND DELINQUENTS PERCEIVE THEIR CHANCES OF APPREHENSION AND CONVICTION FOR THESE HYPOTHETICAL OFFENSES TO BE LESS THAN NONDELINQUENTS PERCEIVE THEIR OWN CHANCES. RESULTS OF THE ANALYSIS OF QUESTIONNAIRE RESPONSES REJECTED THE FIRST HYPOTHESIS, SUPPORTED THE SECOND, AND CONFIRMED THE THIRD IN RELATION TO SELF-PERCEIVED LIKELIHOOD OF BEING ARRESTED. THE DIFFERENCE IN SELF-PERCEPTIONS AS TO PROBABILITY OF CONVICTION IF ARRESTED WAS IN THE SAME DIRECTION AS PROBABILITY OF ARREST, BUT WAS NOT SIGNIFICANT. THE AUTHOR SUGGESTS THAT THE FINDINGS, TAKEN AS A WHOLE, SUPPORT AN EXPLANATION OF THE DIFFERENTIAL IMPACT OF DETERRENCE IN TERMS OF DIFFERENCES IN PERSONALITY EXPRESSED AS PERCEPTIONS OF SELF VIS-A-VIS THE LEGAL STRUCTURE. A TWO-PAGE LIST OF REFERENCES IS PROVIDED. (AUTHOR ABSTRACT MODIFIED)