The results indicate that a changing racial composition contributed to changes in the age composition of Chicago's population, as well as to changes in the homicide rate. Age-specific analyses by race and gender suggest that as much as 24 percent of the total increase in homicide rates arrests and 45 percent of the increase in robbery arrests from 1962 to 1980 could be attributed to an increase in nonwhite men in the population. Increased rates of arrest of nonwhite men appeared to account for large parts of the increases in homicide and robbery arrests, with increased rates for white men and nonwhite women accounting for some of these increases. Data for Chicago and other central urban areas suggest that increases in the 15- to 29-year-old population and increases in the nonwhite male population have an important impact on urban crime trends. 4 tables, 6 figures, and 19 references. (Author abstract modified)