Social science literature links alcohol consumption and human violence and/or crime. Most of the literature in this field only addresses sociological and demographic issues. Crimes occurrences, however, are also location dependent and can be analyzed using GIS methods. Only a handful of studies have analyzed spatial dimensions of crime incidences and their controlling factors. In this paper, we present an in-depth analysis of crime probabilities and their proximity to alcohol-serving establishments based on the locations of 12,458 reported crimes and 247 alcohol-serving establishments in Savannah, Georgia, for the year 2000. Logistic regression and other GIS methods are employed to determine crime density and crime probability in different distance zones from alcohol-serving establishments. The results of our analysis confirm that density and probability of all types of crimes declines exponentially with increasing distance from alcohol services. Current research cannot control for myriad of confounding variables for criminal behavior. Although our findings are consistent with research on alcohol consumption as a causal agent in violent criminal activity, our method of estimating crime probability in space is unique. The crime probability maps produced using this method can be valuable for police resource allocation and patrolling in many other cities.