Making errors or not being able to finish tasks on time is not a rare phenomenon in organizations, especially under the increasing fierce competition and high pressure to finish tasks in a relatively short time. Yet due to motivation of building good image to the supervisors (Leary & Kowalski, 1990), employees are less likely to proactively admit their errors. However, based on the self-cognition theory (Bandura, 1977), it proposed that those with high self-confidence can admit errors. Moreover, this study attempted to explore whether employees’ admitting errors can have detrimental effect of leaders’ evaluation. Overall, this study developed a model that integrates the antecedent as employees' self-efficacy and the outcomes as leaders' evaluation of employees' task and counterproductive behaviors. To examine the causal effect, this study collected data in 2 waves. In time 1, it collected employees’ demographic variables and employees’ self-efficacy; in time 2, it collected employees’ evaluation of their admitting errors behaviors. Also, in time 2, the supervisors were asked to evaluate their subordinates’ performance as task performance and counterproductive behaviors. In short, in time 1, it gained 761 samples, while in time 2, 675 employees and their direct supervisors responded the questionnaire, with a response rate of 89%. In the sample, every 1 leader needs to evaluate 11 employees. To rule out leaders’ impact in evaluating employees’ performance, this study adopted sandwiches estimator from Liu et al.’s (2015) study and used Mplus 6.0 to do the analysis. Results revealed that 1) employees' self-efficacy was positively related to admitting error behaviors (β = .10, p < .05); 2) employees' admitting errors had non-significant relationship with leaders' evaluation of employees' task performance (β = .08, n.s.), but was negatively related to leaders' evaluation of employees' counterproductive behaviors (β = -.19, p < .05). These results demonstrated that admitting errors is not that detrimental to leaders' evaluation of employees' performance, in fact, admitting errors can help to reduce leaders’ evaluation of individuals’ deviant behaviors. This study has both theoretical and practical implications. For the theoretical implications, it figured out the importance of employees’ proactively admitting error behaviors in organizations and illustrated the self-efficacy in influencing employees’ such behaviors; moreover, it depicted that leaders’ evaluations on employees are based more on the information they get rather than just rely on their impressions towards employees, which extends prior studies that put more emphasis on leaders’ impression-based evaluation. In addition, it is also an attempt to explore employees’ moral decisions facing the pro-self vs. pro-organization long-term benefits. For the practical implications, it offered some insights for leaders and employees. Specifically, when behaving in the organizations, employees should make their decisions based on the organizations’ overall benefits, which means sometimes to sacrifice their own interests. Such devotions would be recognized by the leaders, if not being repaid. Moreover, leaders should behave justice and reliable when evaluate employees’ performance, and should rely on facts and information rather than solely on impression. In terms of future studies, it suggested that further studies can explore other antecedents as employees’ conscientiousness, learning goal orientation, integrity, leadership style, organizational learning climate etc. as well as other outcomes like coworkers’ evaluations, the customers’ evaluation, to enrich the studies on employees’ proactively admitting errors in organizations.