The data came from the second phase of a 1998 Department of Housing and Urban Development study that used geographic information systems to measure levels of crime in and around public housing developments. Both cities were in the eastern United States and had populations in the hundreds of thousands. Crime data came from local police agencies for the FBI's Part I crimes (i.e., homicide, rape, aggravated assault, robbery, burglary, larceny, auto theft, and arson) for February through July 1998. Additional data came from computerized crime mapping. Results in both cities revealed that black female residents of public housing were at a much higher risk of aggravated assault than were black and white women who lived elsewhere in the same jurisdiction. However, the geographic pattern for aggravated assaults for black female public housing residents differed markedly in the two cities. The analysis used the perspective of situational crime prevention to attribute the differences in victimization patterns to the different architectural design and geographic dispersion of the respective cities’ public housing developments. The analysis revealed that one city’s developments offered less privacy and accessibility than did those of the other city, thus discouraging would-be assailants. Tables, notes, and 39 references (Author abstract modified)