THIS IS REFLECTED IN THE THEORY THAT, GIVEN A CITY THAT CONTAINS ONLY A SMALL PROPORTION OF THE RESIDENTS IN THE LARGER ECOLOGICAL COMMUNITY, THE CONVENTIONAL CRIME RATE FOR THAT CITY COULD BE HIGH MERELY BECAUSE THE DENOMINATOR OF THE RATE UNDERESTIMATES THE POTENTIAL NUMBER OF VICTIMS OR OFFENDERS. THE RELATIONSHIP DOES HOLD FOR MANY AMERICAN CITIES WHEN URBANIZED AREAS (UAS) OR STANDARD METROPOLITAN STATISTICAL AREAS (SMASS) ARE TAKEN AS APPROXIMATIONS OF COMMUNITIES; BUT IT HOLDS ONLY FOR SINGULAR CITIES, EACH OF WHICH IS THE ONLY CENTRAL CITY IN A SMSA. THE ARGUMENT OF THIS PAPER IS THAT SINGULAR CITIES ARE MUCH MORE HOMOGENEOUS AS REGARDS DOMINANCE WITHIN THE COMMUNITY THAN ARE OTHER TYPES OF CITIES, AND THAT DOMINANCE DETERMINES THE EXTENT TO WHICH A CITY WILL ATTRACT NONRESIDENT PARTICIPANTS IN CRIMES. IN ANY CASE, THE FINDINGS CAST DOUBTS ON THE USE OF CONVENTIONAL CRIME RATES FOR CITIES IN TESTING THEORIES. REFERENCES ARE INCLUDED. (AUTHOR ABSTRACT)