The decision to withhold adjudication was used as an indicator of sentencing severity. The data included 4,148 of the 4,246 felony convictions during the study period; the small number of cases involving first or second-degree murder, forcible rape, or kidnapping were excluded, because most or all these offenders would be ineligible for an adjudication withheld sentence. The sample included 2,231 cases with black offenders, 975 cases with Hispanic offenders, and 942 cases with white offenders. Results provided only partial support for the study hypothesis. Race/ethnicity interacted with gender and with the offender's prior criminal record to influence the decision to withhold adjudication. Among offenders with no prior felony convictions, neither race nor gender was a significant predictor of the decision. In contrast, Hispanic male repeat offenders and females repeat offenders in all three racial/ethnic groups were more likely than black male repeat offenders to have adjudication withheld. Further analysis revealed that the effect of gender was confined to offenders with a prior conviction for a nonviolent offense. Findings highlighted the importance of using an interactive model that incorporates the effects of race, ethnicity, and gender. Further research should continue the effort to determine the conditions under which race and gender affect sentence severity. Tables, notes, and 57 references (Author abstract modified)