The study found that changes in misdemeanor arrest rates were negatively related to changes in total homicide rates. This supports Keeling and Ceases general claim that the implementation of new policing policies (order-maintenance policing) had a noticeable role in the homicide decline in New York City. The study also found that when homicides were disaggregated by weapon use, the effect of changes in policing were significant for gun-related, but not for nongun-related homicide. Changes in homicide rates were also significantly affected by changes in cocaine use. Declining cocaine use, as indicated by the toxicology of accident victims, was associated with decreasing homicide levels. This study went beyond previous research in this area by examining gun-related and nongun-related homicides separately, by using a superior measure of drug activity based on cocaine toxicology of accident fatalities, and by including a measure of felony arrests in order to help distinguish the effects of increased policing of disorder offenses from intensified policing generally. Analyses were based on pooled, cross-sectional, time-series data for 74 New York City precincts over the 1990-99 period, with all variables measured at the appropriate precinct level. 3 tables, 1 figure, 45 references, and appended univariate statistics for New York City Police Precincts, 1990-99