The analysis was based on 47 comparisons that involved a total of 48,251 cases. Regression coefficient values from each independent study were separated into the categories of the effect size of race, prior criminal record, and the type/severity of the offense. Each regression coefficient was standardized to an effect size estimate, converted to a z score, and then weighted for sample size. Results indicated that although the size of the effect of race on sentencing is not statistically significant at this level of aggregation, the measurement techniques used by researchers significantly condition the relationship of race to sentencing. Specifically, certain methods of classifying racial groups may mask the true effect of race on sentencing decisions. Thus, although the severity of the current offense was the most significant predictor of sentence length, the true effect of race on sentencing outcomes may be hidden behind differences in how researchers specify the race variable. Findings indicated that the increased formalization, professionalization, and routinization of the judiciary, as well as the proliferation of sentencing guidelines and determinate sentencing laws, have resulted in a situation in which sentencing decisions tend to be based largely on legally relevant factors only. Nevertheless, future research on differential handling of racial minorities should give special attention to measurement techniques and the impact of choosing a particular racial classification scheme. Table, notes, and 88 references (Author abstract modified)