To overcome methodological limitations, data included official records and self-reports from 1,400 inmates. Data came from the California Offender-Based Transaction Statistics for 1980 and the Rand Inmate Survey. Multiple regression analyses were used to control for factors other than race. Although the case processing system usually treated offenders similarly, minority suspects were more likely to be released after arrest and were more likely to be given longer sentences in prisons (as opposed to jails) following a felony conviction. No significant differences between white and minority inmates were found with respect to participation in treatment and work programs. In California and Texas, minority inmates served longer sentences than did whites, while in Michigan, blacks received longer sentences but served roughly the same time as whites. There were racial differences in criminal motivation, weapons use, and prison behavior, but most were not statistically significant. In the case of greater postarrest release favoring minorities, factors such as defendant lack of cooperation, differences in arrest procedures, and evidentiary problems may have contributed to the findings. In general, if discrimination existed, it was inconsistent. While minorities were given longer, harsher sentences and served longer sentences in two of three States studied, this may result from differences in who makes decisions and what information is used and how it is weighted. Tabular data, one note, and seven references are included.