The theory of drug-offender sentencing developed and tested draws on two bodies of theory in the literature on sentencing and racial disparity. The first, called the "racial stereotypes" approach, emphasizes the importance of racial stereotypes in the assessment of individual offenders. The second, called the "case processing" approach, focuses less on racial-ethnic disparities than on routine processing criteria set in an organizational context. The theory was tested by examining all sentences ordered for Black and White offenders convicted of felony drug offenses between July 1, 1995, and December 31, 1998, in Washington State (n=22,858). The dependent variables were "incarceration," as measured by whether an individual was sentenced to more than 30 days of incarceration, and sentence length, as measured by term of confinement in months. Independent variables were being male, being a dealer, and having a prior felony record. The study found that White offenders who most closely resembled the stereotype of a "dangerous" drug offender (being a male drug dealer with a prior felony record) received significantly harsher sentences than other White drug offenders who did not fit this stereotype. On the other hand, being a "dealer" had a significantly larger effect on the likelihood of incarceration for White offenders than for Black offenders. Being both a dealer and having a prior felony record also had significantly larger effects on sentence length for White offenders than for Black offenders. In the less serious categories of offenders (nondealers and male nondealers with no priors), however, judges were less likely to incarcerate White offenders than their Black counterparts. 6 tables and 39 references