The study found that gender, age, and employment status, but not race/ethnicity, had direct effects on sentencing. These findings are conditioned by the fact that an offender's gender and employment status had effects on sentencing only in cases sentenced outside the Federal sentencing guidelines. The effect of gender and employment status were conditioned by race/ethnicity. No support was found for the hypotheses that young Black and Hispanic men and unemployed Black and Hispanic men would receive more severe sentences. Both Black and Hispanic women, however, received substantially shorter sentences than their male counterparts. Being employed benefitted White offenders at sentencing. A number of case characteristics had differential effects on the sentences imposed on White, Black, and Hispanic offenders. Pretrial status, for example, had a significant effect on sentence length for Black offenders and White offenders, but not for Hispanic offenders. Compared to their counterparts who were released prior to sentencing, Black offenders who were in custody received sentences that were a year and a half longer than offenders not in custody. Black offenders and Hispanic offenders were more likely than White offenders to be held in custody prior to sentencing. Pleading guilty produced more lenient sentences for Black offenders and for White offenders, but not for Hispanic offenders. The data were collected from three U.S. district courts in the Eighth Circuit: the District of Minnesota, the District of Nebraska, and the Southern District of Iowa. All offenders sentenced for a drug offense (major charge) in fiscal years 1998, 1999, and 2000 (n=1,752) were included in the study. 7 tables and 49 references