ABSTRACT
Questions regarding the quality and comparability of cross-national crime data from different sources have cast doubt on the validity of explanatory cross-national research on crime. While past research has addressed some of the important issues, other problems remain that have not been adequately addressed Data sources differ in the types of crime for which rates are available, rates for the same nation for the same type of crime, and the sample composition of nations for which rates are available. This research examines whether these differences affect explanatory cross- national research on crime in terms of the association of relevant independent variables with crime rates. Results show that choice of data source affects the associations of relevant independent variables to homicide rates and that these associations also differ depending on whether robbery or other types of theft is studied Recommendations are made on choice of rates and data sources for future research.