This research seeks to add a new dimension to the growing body of literature investigating disparities in sentencing outcomes. Our study used multilevel modeling techniques to incorporate legal and extralegal individual-level variables with macro-level data in an examination of the factors influencing the formal application of the criminal label when identical noncriminal sanctions are available. Results suggest that measures of police racial profiling are predictive of unfavorable plea outcomes and subsequent application of a criminal label. Findings indicated that the likelihood of receiving an unfavorable criminal plea bargain was greater for men, non-Whites, defendants with a prior arrest, defendants arrested in areas of low concentrated disadvantage, and defendants arrested in areas with high levels of racial profiling. In addition, both concentrated disadvantage and stop disparity conditioned the effect of race on plea outcome.