Highlighting resource inequality, social processes, and spatial interdepend
ence, this study combines structural characteristics from the 1990 census w
ith a survey of 8,872 Chicago residents in 1995 to predict homicide variati
ons in 1996-1998 across 343 neighborhoods. Spatial proximity to homicide is
strongly related to increased homicide rates, adjusting for internal neigh
borhood characteristics and prior homicide. Concentrated disadvantage and l
ow collective efficacy-defined as the linkage of social control and cohesio
n-also independently predict increased homicide. Local organizations, volun
tary associations, and friend/kinship networks appear to be important only
insofar as they promote the collective efficacy of residents in achieving s
ocial control and cohesion. Spatial dynamics coupled with neighborhood ineq
ualities in social and economic capacity are therefore consequential for ex
plaining urban violence.