If discrimination against an historically oppressed social group is dismantled, will the group forge ahead? The authors present experimental evidence that a history of social and legal disabilities may have persistent effects on a group's earnings through its impact on individuals' expectations. In the first experiment, 321 high-caste and 321 low-caste junior high school male student volunteers in rural India performed the task of solving mazes under economic incentives. There were no caste differences in performance when caste was not publicly revealed, but making caste salient created a large and robust caste gap. When a nonhuman factor influencing rewards (a random draw) was introduced, the caste gap disappeared. To test whether the low caste's anticipation of prejudicial treatment caused the caste gap, the authors conducted a second experiment that manipulated the scope for discretion in rewarding performance. When the link between performance and payoffs was purely mechanical, making caste salient did not affect behavior. Instead, it was in the case where there was scope for discretion and judgment in rewarding performance that making caste salient had an effect. The results suggest that when caste identity is salient, low-caste subjects expect that others will judge them prejudicially. Mistrust undermines motivation. The experimental design enables the authors to exclude as explanations of the caste gap in performance socioeconomic differences and a lack of self-confidence by low-caste participants.