Companies in the service industry have greater pressure now to improve the quality of their customer service. Employees in service organizations usually should display an appropriate emotion toward their customers. Previous studies on emotional labor strategies (surface acting and deep acting) examined mainly their impacts on employees’ attitudes and cognition. Till now, the impacts of emotional labor strategies on employee sabotage behavior have not been studied thoroughly yet. Moreover, insufficient attention has been paid on understanding the boundary conditions that may enhance or mitigate the effects of emotional labor strategies on employees’ sabotage behavior. Therefore, based on the conservation of resources theory and the emotion theory, this study attempted to examine the impacts of emotional labor strategies on employees’ sabotage behavior, as well as on the moderating roles of policy strength perception and the social sharing of emotions. Data was collected from employees working in call centers of two companies in Beijing, China. Two waves of the survey were conducted. In the first wave, participants were required to complete a questionnaire including demographic information (e.g., gender, age, and education), control variables (neuroticism and negative affectivity), emotional labor strategies (surface acting and deep acting), policy strength perception, and social sharing of emotions. In the second wave, participants reported the levels of service sabotage behavior. We invited 1014 employees to participate in the first wave of the survey, and 899 employees in the second wave. The final sample consisted of 788 employees who completed both two waves. We examined our hypotheses with the SPSS 18.0 software. Results showed that: 1) surface acting was positively related to employee sabotage behavior, but deep acting was not significantly associated with employee sabotage behavior. 2) The relationship between surface acting and employee sabotage behavior was significantly moderated by policy strength perception. Specifically, surface acting significantly enhanced employee sabotage behavior when the level of policy strength perception was high. 3) Social sharing of emotions significantly moderated the relationship of deep acting and employee sabotage behavior. Specifically, when the level of social sharing of emotions was high, deep acting significantly reduced employee sabotage behavior. This study contributes to the conservation of resources theory and emotional labor literature in several aspects. First, we demonstrate the impacts of different emotional labor strategies on employee sabotage behavior. It significantly extends our understanding of the behavioral outcomes of emotional labor strategies. Second, we highlight the value of conservation of resources theory in the emotional labor work context by demonstrating two important boundary conditions that either enhance or mitigate the impacts of emotional behavior strategies on employee sabotage behavior. Accordingly, managerial implications regarding the alleviating role of policy strength perception and social sharing of emotions in emotional labor management are discussed. We advocate that service companies should provide special training programs for employees to learn ways to cope with emotional labor activities. They should establish clear behavioral norms and reward systems as guidance for employees. In addition, they could provide employees opportunities, such as employee assistance programs, to communicate and share with others.