Siting of drug and alcohol treatment facilities is often met with negative reactions because of the assumption that these facilities increase crime by attracting drug users (and possibly dealers) to an area. This assumption, however, rests on weak empirical footings that have not been subjected to strong empirical analyses. Using census block groups from Philadelphia, PA, it was found that the criminogenic impact of treatment facilities in and near a neighborhood on its violent and property crime rates may be contingent on the socioeconomic status (SES) of the neighborhood. Paying attention to both the density and proximity of facilities in and around neighborhoods, results showed that the criminogenic impact of treatment facilities depended largely on neighborhood SES. Under some conditions, more treatment facilities nearby were associated with lower crime. Reasons, why the presumed criminogenic impact of treatment facilities appears only under some conditions, were suggested.