This study integrates research on social networks and interpersonal counterproductive behaviours to examine the role of third party relationships in predicting an individual's susceptibility to co-worker mistreatment, and in moderating the relationship between co-worker mistreatment and job performance. Third party embeddedness and network closure are examined in the formal workflow network and the informal liking network. Results obtained from employees in a family-owned Chinese business in Singapore indicate that an individual is more likely to be mistreated by a co-worker when both parties are strongly embedded in mutual third party relationships in the workflow network, and that the individual is less likely to be mistreated when both parties are strongly embedded in the liking network. At the individual network level, network closure (i.e., the extent to which an individual's contacts are themselves connected to one another) in the workflow network increases the likelihood that a co-worker will mistreat the individual, but closure in the liking network weakens the negative relationship between mistreatment and performance. The findings offer a network-based perspective to understanding interpersonal mistreatment and counterproductive work behaviours, particularly in the context of Confucian Asian firms, and provide practical implications for organizations and individuals to reduce counterproductive behaviours at work.