This work is under a CC-BY-NC-SA license. You are free to copy and redistribute the material in any format as well as remix, transform, and build upon the material as long as you give appropriate credit to the original creator, provide a link to the license, and indicate any changes made. You may not use this work for commercial purpose and must distribute any contributions under an identical license.Interpersonal confrontation has been heralded in the stereotyping and prejudice literature as a situationally flexible, personally empowering, and highly effective prejudice reduction approach (Czopp & Ashburn-Nardo, 2012; Mallett & Monteith, 2019b). Indeed, a number of experiments consistently show that confrontation (compared to ‘no confrontation’) reduces confrontees’ stereotyping and prejudice endorsement, even among high-prejudice confrontees who reject egalitarian values (e.g., Burns & Monteith, 2018; Chaney & Sanchez, 2018; Czopp et al., 2006). These experiments, however, have uniformly tested confrontation efficacy in social settings where egalitarian norms are strong. This is problematic because norm compliance pressure is theorized to be a key mechanism explaining how confrontation regulates prejudice expression (Czopp et al., 2006). The present research addresses this limitation by comparing confrontation effectiveness across situations where prejudice expression is deemed socially acceptable and unacceptable. In Study 1, college students’ forecast how they would feel, think, and behave in response to being confronted. In Studies 2 and 3, college students’ biased responses were confronted and they reported their feelings, thoughts, and behavioral intentions, while their subsequent stereotyping behaviors were unobtrusively measured. A stark divide between forecaster expectations and experiencer realities emerged. Study 1 analyses revealed that in situations where prejudice acceptability was high (vs. low and moderate), college student forecasters expected prejudice expression to be less offensive and, consequently, anticipated feeling less guilty for expressing prejudice and weaker motivation to self-correct. These same forecasters also anticipated that confrontation would make them feel angrier and expected to express more dismissiveness and hostility. In Studies 2 and 3, however, self-corrective reactions to confrontation emerged regardless of prejudice acceptability level. College students confronted for expressing prejudice (vs. not confronted) rated their own stereotyping behaviors as more offensive. In turn, they reported feeling more guilt and a stronger desire to self-correct. Additionally, these students reduced their degree of behavioral stereotyping following the confrontation. Taken together, these findings suggest that, despite forecaster intuitions, confrontation can be an effective prejudice reduction tool, even in situations where prejudice expression is widely considered socially acceptable. Theoretical and applied implications of this work are discussed, as well as directions for future research.