The researchers studied official police statistics regarding crime in 29 conventional housing developments with almost 35,000 residents, accounting for more than half of the entire public housing population within each city. An address matching procedure was used for nearby census tracts to determine offense rates in areas of private housing. Results revealed clearly that drug and violent offenses are severe problems in housing developments. From 1986 to 1989, average annual rates of drug law offenses in housing developments were 33 per 1,000 residents in Washington, D.C., 53 per 1,000 in Phoenix, and 58 per 1,000 in Los Angeles. Rates of violent offenses were even higher: 41, 54, and 67 in Washington, Phoenix, and Los Angeles respectively. In all three cities, these rates are higher than citywide or nearby neighborhood rates for the same kinds of offenses. In contrast, property offense rates in Los Angeles and Washington, D.C. are considerably lower than citywide rates. Results also confirmed that some housing developments had much more severe crime problems than others; some had rates lower than citywide rates. In addition, police activity in housing developments is roughly proportional to the offense rates. Findings indicated that it is reasonable to devote a disproportionate share of drug and law enforcement resources to public housing developments and that crime control initiatives should focus tightly on the problems of particular developments. Additional findings and implications, figures, tables, footnotes, appended methodological information and additional results, and 109 references