This empirical study investigates the influence of ambivalent supervision style on employee outcomes. Ambivalent supervision style is a new form of leadership, which is categorized as a form of destructive leadership; however, there is scant literature on ambivalent leadership in organizational studies. Numerous studies have pointed out that there is a need to study this form of leadership and its influence on employee outcomes. Thus, this research is an attempt to understand the phenomenon of influence of ambivalent supervision style on positive and negative employee outcomes. It has been hypothesized that ambivalent leadership has significant negative effect on employees’ outcomes of task performance, organizational citizenship behaviors directed towards individual and organization and positive influence on employees’ counterproductive work behaviors directed towards individual and organization. Additionally, employee state paranoia as underlying psychological process has been investigated as a mediating variable and co-worker support as a buffering mechanism. By doing so, the current research advances the role of employee state paranoia as a mediating mechanism. The framework of the study is guided by transactional theory of stress, which proposes that employees engage in coping mechanisms to offset the effects of stressors. Thus, co-worker support as a coping mechanism has shown positive trends. The study used quantitative research design and cross sectional data were collected through self-administered questionnaires distributed among the service sector organizations in Punjab, Pakistan. A total of 645 respondents corresponding to 129 dyads were found usable for data analysis. Confirmatory factor analysis and multi-level Hierarchical Linear Modeling was used to test the moderated mediation model. Results support the hypothesized relationships. The results indicate that ambivalent supervision style negatively effects positive employee outcomes and positively effects employee negative outcomes (counterproductive work behaviors). The results further highlight that employee state paranoia partially mediates the relationship between ambivalent supervision style and employee outcomes. The buffering mechanism using coworker support holds true for positive employee outcomes but is insignificant for negative employee outcomes.