We analyzed the relationship between crime and indicators of residential yard management in Baltimore City and County. Data came from a survey we conducted of over one thousand front yards that included more than 40 indicators relating to lawns, trees, shrubs, beds and other features. These indicators were related to point counts of crime at the 150 m scale using a combination of ordinary least squares, spatial error, and Poisson regressions. After controlling for income, population density, block-scale tree canopy, and housing type, we found a consistently significant relationship between crime and a number of indicators of yard management. Yard-level variables that were negatively associated with crime included: the presence of yard trees, garden hoses/sprinklers, and lawns, in addition to the percentage of pervious area in a yard. Those positively associated with crime included presence of litter, desiccation of the lawn, lack of cutting of the lawn, and number of small trees in front of or adjacent to the property. While these results do not establish causality, they add evidence to a growing literature that suggests the possibility of several mechanisms by which environmental design may reduce crime: ?cues to care? (the inverse of the ?broken window? hypothesis) can lead to reduced crime by signaling to criminals the presence of social capital and the active involvement of neighbors in community spaces; and more appealing landscaping draws more ?eyes on the street,? which in turn deters criminals.