The study focused on the black-white transition in neighborhood racial composition. The residential city blocks of Cleveland were the units of analysis. Dependent variables were the frequencies of crimes per block and the changes in the number of incidents per block over the decade. Three sets of independent variables were used to represent the social composition, residential environment, and other demographic characteristics of the blocks which could affect the incidence of crime. Data analyses indicate that while change in racial composition of neighborhoods from white to black had relatively powerful standardized effects on 1980 crime rates in regressions using only changes in block characteristics, these diminished considerably once the effects of past block characteristics were controlled. Although racial characteristics of blocks retain their significance for all crimes except rape and grand theft, the effects are much smaller in size than those for other variables. Variables of greater significance included residents on the block in 1970, 1970 crime rates, concentration of primary individuals on the block, and change in number of residents on the blocks. Thus racial change had only small effects on changes in crime levels, which derive more from historical persistence and changes in crime over the decade. 6 tables and 38 references. (Author abstract modified)