AbstractTo test the assumption that police reports cannot be used as a representative sample of all criminal events, a research study was undertaken comparing the characteristics of robberies and robbery victims derived from police offense reports in San Jose, California, to those derived from a victimization survey in the same area The result of the comparison of these two data sources disclosed that: (1) for commercial robberies there were no differences between police offense reports and victimization survey data for the characteristics which were compared; (2) for personal robberies, there were no differences for the characteristics of victim's sex and age, number of victims, victim/offender relationship, and number of offenders. The differences observed for victim's race and time of occurrence can be attributed to classification problems and methodological differences It is highly probable that the police robbery offense reports and the Victimization survey data are representative of the same population of robberies Results suggest that further evaluation of the assumption that police offense reports do not reflect the “true” nature of all (i.e., reported and unreported) crime is warran