DATA FOR 656 UNITED STATES CITIES WITH A 1960 POPULATION OF 25,000 OR MORE WERE ANALYZED. CITIES HAVING A POPULATION OF 100,000 OR MORE WERE CONSIDERED SEPARATELY FROM SMALLER COMMUNITIES. THE DEPENDENT VARIABLES WERE 1967 RATES FOR EIGHT MAJOR CRIMES - MURDER, NONNEGLIGENT MANSLAUGHTER, FORCIBLE RAPE, AGGRAVATED ASSAULT, ROBBERY, BURGLARY, AUTO THEFT, AND LARCENY OF $50 OR MORE. THE EFFECT OF CROWDING ON EACH CRIME RATE WAS ASSESSED THROUGH MULTIPLE AGGRESSION. OVERALL, IN LARGE CITIES, CROWDING WAS FOUND TO BE RELATED TO A VARIETY OF CRIME RATES, EVEN AFTER CONTROLLING FOR VARIABLES SUCH AS REGION, RACE, EDUCATION, INCOME, AGE, POPULATION SIZE, AND NATIVITY. AREAL CROWDING ACCOUNTED FOR 3% WHILE HOUSEHOLD CROWDING ACCOUNTED FOR LESS THAN 1% OF THE EXPLAINED VARIATION, AND APPEARED ONLY IN CITIES WITH A HIGH DWELLING UNIT CONCENTRATION. AREAL DENSITY HAD MORE RELATIONSHIP WITH PROPERTY CRIME THAN WITH CRIMES AGAINST PERSONS. CROWDING EXPLAINED LESS VARIATION IN CRIME RATES IN THE SMALLER CITIES, ALTHOUGH HOUSEHOLD CROWDING WAS MARKEDLY RELATED TO SIX OF THE EIGHT CRIMES. REFERENCES ARE INCLUDED. FOR A COMMENT ON THIS STUDY AND A REPLY TO THE COMMENT, SEE NCJ-036638 AND 036743. (AUTHOR ABSTRACT MODIFIED)