This paper examines problems in studying the relationship between casinos and crime, with a focus on a recently published, influential study (Grinols and Mustard 2006) which concluded that casinos cause a significant amount of county-level crime in the U.S. Five key issues are examined. First, the most serious problem with their analysis is that it uses a crime rate that excludes the visiting population at risk, thereby overstating the crime rate in casino counties. Second, the crime data used are potentially inaccurate. Third, the results may suffer from a bias caused by counties self-selecting into the “casino county” category. Fourth, the dummy variables used to account for casinos do not allow the authors to isolate the crime effect caused by casinos. Finally, the authors make conclusions that are not supported by their data, analysis, and results. An examination of these issues is important because it will shed additional light on the debate over the effects of casinos, and provides valuable information for subsequent researchers who study the casino-crime relationship.I am grateful to Jay Albanese, Bill Eadington, David Forrest, Mark Nichols, Don Ross, and Richard Thalheimer for helpful comments and suggestions, and especially to John Jackson and Ben Scafidi for helpful discussions on this paper. I am responsible for the content and any errors.