PREVIOUS RESEARCH HAD INDICATED THAT TELEVISION VIEWING CAUSES PEOPLE TO OVERESTIMATE THE AMOUNT OF DANGER IN THEIR OWN NEIGHBORHOOD, BUT HAD NOT CONTROLLED FOR THE ACTUAL INCIDENCE OF CRIME IN THE NEIGHBORHOOD. THE FOUR NEIGHBORHOODS SURVEYED INCLUDED HIGH-CRIME AND LOW-CRIME AREAS IN BOTH DOWNTOWN AND SUBURBAN TORONTO. IN THE PRESENT DOOR TO DOOR SURVEY, BETWEEN 69 AND 83 PEOPLE IN EACH GEOGRAPHIC AREA WERE ASKED TO COMPLETE A 37-ITEM FIXED ALTERNATIVES QUESTIONNAIRE. BOTH HOUSEHOLDS AND HOUSEHOLD MEMBERS WERE SELECTED RANDOMLY. THE NUMBER OF PROGRAMS WATCHED WAS USED AS AN INDEX OF TOTAL VIEWING; PROGRAMS WERE DIVIDED INTO VIOLENT AND NONVIOLENT TYPES BEFORE FURTHER DATA TABULATION. MULTIPLE REGRESSION ANALYSIS AND CANONICAL CORRELATION ANALYSIS PRODUCED SIMILAR RESULTS. WHEN THE FOUR AREAS WERE POOLED, THE RESULTS REPLICATED PREVIOUS FINDINGS. IN CONTRAST, THE AVERAGE CORRELATIONS WITHIN EACH GEOGRAPHIC AREA WERE INSIGNIFICANT. THIS SUGGESTS THAT WHEN ACTUAL CRIME INCIDENCE IS CONTROLLED FOR, THERE IS NO OVERALL RELATIONSHIP BETWEEN TELEVISION VIEWING AND FEAR OF BEING A CRIME VICTIM. A DESCRIPTION OF THE METHODOLOGY, TABLES INCLUDING BOTH RESULTS AND EXAMPLES OF THE QUESTIONS USED, A FOOTNOTE, AND REFERENCES ARE INCLUDED. (CFW)