Expanding on conflicting theoretical conceptualizations of implicit bias, 6 studies tested the effectiveness of different procedures to increase acknowledgment of harboring biases against minorities. Participants who predicted their responses toward pictures of various minority groups on future implicit association tests (IATs) showed increased alignment between implicit and explicit preferences (Studies 1–3), greater levels of explicit bias (Studies 1–3), and increased self-reported acknowledgment of being racially biased (Studies 4–6). In all studies, effects of IAT score prediction were significant even when participants did not actually complete IATs. Effects of predicting IAT scores were moderated by nonprejudicial goals, in that IAT score prediction increased acknowledgment of bias for participants with strong nonprejudicial goals, but not for participants with weak nonprejudicial goals (Study 4). Mere completion of IATs and feedback on IAT performance had inconsistent effects across studies and criterion measures. Instructions to attend to one’s spontaneous affective reactions toward minority group members increased acknowledgment of bias to the same extent as IAT score prediction (Study 6). The findings are consistent with conceptualizations suggesting that (a) implicit evaluations are consciously experienced as spontaneous affective reactions and (b) directing people’s attention to their spontaneous affective reactions can increase acknowledgment of bias. Implications for theoretical conceptualizations of implicit bias and interventions that aim to reduce discrimination via increased acknowledgment of bias are discussed.