National victimization survey data for the years 1972 to 1976 are examined in order to investigate the extent to which they are in accord with Uniform Crime Report arrest data regarding the offender's sex. The results for common-law crimes–both the personal crimes of rape, robbery, assault, and personal larceny, as well as the household crimes of burglary, larceny, and motor vehicle theft–parallel arrest data in showing that male involvement in these crimes is proportionately much greater than is the involvement of females. Except for a slight recent increase in female involvement for larceny, there did not appear to be systematic changes in the proportionate involvement of females across the few years of victimization data now available. Victims of female-offender crimes reported them less often to the police than was true for victims of male-offender crimes. But the latter crimes are usually more serious and male-offenders dominate commercial robbery, almost always reported to the police. The chivalry hypothesis–that male victims of female offenders are reluctant to report female-offender crimes to the police–was not supported. Even with seriousness of victimizations controlled, males tended to report their victimizations to the police more when they were victimized by females; the opposite was true for female victims. Overall, the results are consistent with the argument that sex is an important nonartifactual correlate of involvement in criminal behavior. Theories that fail to recognize and accommodate the importance of sex–and other traditional demographic correlates (e.g., race) of involvement in common-law crimes–are ignoring important factors accounting for criminal activity.