The data represented a stratified random sample of females convicted of assault, petty theft, forgery, and drug law violations and listed in a northern California probation department's files between 1971 and 1976. A substantial discrepancy was found between the offenses for which the women were arrested and those for which they were convicted. The visibility of the conduct involved, the legal system's priorities in prosecuting cases, and plea bargaining may all have contributed to this discrepancy. Analysis of conviction records showed that diverse behaviors were included in each of the offense categories except for petty theft. Whereas the petty theft offenders were generally young and committed their offenses against merchants, those convicted of other crimes had engaged in actions varying widely in seriousness and were sometimes accessories to others' actions. Assault victims were commonly known to the offenders, while forgery victims were generally strangers. Black women were more commonly convicted of crimes of violence, Hispanic American women of burglary, and white and Native American woman of fraud and forgery. Other aspects of women's social characteristics also appeared to be influential in the sanctioning decision. Official crime statistics are misleading in that they overlook many qualities of the crimes involved. Data tables, notes, and a list of 40 references are supplied.