Sentencing reforms that seek to reduce racial disparities in punishment, either on equity or efficiency grounds, are likely to fail if they replicate racial sentencing patterns of the past. Further, the widespread use of indeterminate sentences in the United States has resulted in discriminatory sentencing. Evidence of whether sentencing reforms have narrowed racial disparities in incarceration rates must be examined in the larger context of efficiency gains or losses associated with reforms intended to provide equity through reductions in sentencing disparities. Data from the U.S. Board of Parole on racial disparities in parole release rates during the 1970's demonstrate that original parole guidelines led to racially discriminatory parole release practices, in the sense that factors such as marriage and prior prison sentences had differential effects on the release rates of black and white inmates. These data also showed that differentials could not be explained by differences in observed characteristics of black and white inmates. The role of the U.S. Sentencing Commission in monitoring the implementation of sentencing guidelines is discussed. 55 footnotes and 3 tables