CRIME RATES FOR 6 OFFENSES IN 10 MINNEAPOLIS NEIGHBORHOODS WERE CALCULATED AND COMPARED TO SURVEY RESPONSES OF 1,541 ADULT RESIDENTS OF THE SAME NEIGHBORHOODS. SURVEY RESPONDENTS ASSESSED THE EXTENT OF THE CRIME PROBLEM IN THEIR NEIGHBORHOODS, THE LIKELIHOOD THAT THEY THEMSELVES WOULD BE VICTIMIZED, AND THE DANGEROUSNESS OF WALKING THROUGH THEIR NEIGHBORHOODS ALONE AT NIGHT. THE CITIZENS' PERCEPTIONS OF THE SERIOUSNESS OF THE CRIME PROBLEM IN THEIR NEIGHBORHOODS WERE QUITE ACCURATE. FOR CRIMES MOST LIKELY TO GENERATE FEAR--ROBBERY, ASSAULT, RAPE--CITIZENS' ESTIMATES OF VICTIMIZATION PROBABILITIES WERE SIGNIFICANTLY RELATED TO ACTUAL RATES. RATES WERE ALSO SIGNIFICANTLY CORRELATED WITH CITIZENS' PERCEPTIONS OF THE DANGER INVOLVED IN BEING OUT ALONE AFTER DARK. THE FINDINGS IMPLY THAT IT WOULD BE A MISTAKE TO MAKE GENERALIZATIONS ABOUT CITIZENS' ATTITUDES TOWARD CRIME FROM NATIONAL-LEVEL DATA, WHICH HAVE SHOWN THAT PEOPLE TEND TO EXAGGERATE THE CRIME PROBLEM AND TO BE UNJUSTIFIABLY FEARFUL. THE FINDINGS ALSO IMPLY THAT POLICIES DESIGNED TO REDUCE INCIDENCE OF CRIME WILL ALSO REDUCE THE FEAR OF CRIME. SUPPORTING DATA AND A LIST OF REFERENCES ARE INCLUDED. (LKM)