Guidelines sentencing data from Pennsylvania for the years 1985–1987 are analyzed to assess the influence of gender on judges' imprisonment decisions. These data provide detailed information on offense severity and prior record, permit statistical controls for other variables thought to affect imprisonment decisions, cover a fairly comprehensive list of common-law offenses (with adequate sample size), and contain judges' dispositional-departure reasons for sentences outside the guidelines schema. The data—analyzed with additive and interactive models–indicate that gender (net of other factors) has a small effect on the likelihood of imprisonment toward lesser jailing of female defendants but has a negligible effect on the length-of-imprisonment decision. Observations and interview responses from selected judges help to clarify the ways in which judges' sentencing practices are gender linked. Together, the statistical and the qualitative data suggest that the sentencing practices of judges are driven by two main concerns, blameworthiness (e.g., as indicated by prior record, type of involvement, remorse) and practicality (e.g., as indicated by child-care responsibility, pregnancy, emotional or physical problems, availability of adequate jail space). Based on our findings, we suspect that when men and women appear in (contemporary) criminal court in similar circumstances and are charged with similar offenses, they receive similar treatment. A major question from a policy perspective is, when gender disparities in sentence outcomes do arise, are the disparities warranted or unwarranted?