Macro-level research has established that a variety of community traits are likely to influence burglary rates across neighborhoods. However, these studies have almost never considered the effects of aggregate alarm ownership despite some evidence of the importance of community-level target hardening. Thus, the present study examined the relationships among tract-level alarm ownership, burglary rates, and sociodemographic variables using a sample of 90 census tracts from a large city in the Northeastern United States. Regression models indicated an unexpected positive association between alarm ownership and burglary rates, although that relationship may be moderated by poverty rates. Additionally, the strength of the association between burglary and alarm ownership appeared to vary widely from year to year.