The open innovation paradigm emphasizes that firms can improve their innovation performance by collaborating with other firms. However, there is no consensus in the open innovation literature regarding what drives firms' implementation of inbound and outbound open innovation and how the two types of open innovation influence innovation and firm performance. The present meta‐analysis synthesizes previous research to advance our understanding of open innovation by reexamining its antecedents and its moderating effects on innovation and firm performance. Using data from 157 articles, our results show that appropriation strategy, formal organizational structure, learning culture and top manager competence drive inbound open innovation, while innovation strategy, appropriation strategy, learning culture, and top manager competence drive outbound open innovation. In addition, we find that absorptive capacity positively moderates the positive impact of inbound and outbound open innovation on innovation performance. However, industry type (i.e., high‐ and low‐tech) is not found to moderate the effect of open innovation on innovation or firm performance, suggesting that both high‐ and low‐tech firms can equally reap benefits from open innovation. Furthermore, this study also partially supports the moderating effects of firm size and cultural value (individualism vs. collectivism) on the open innovation—performance relationship. These findings collectively suggest that the effects of inbound and outbound open innovation on performance are subject to contextual factors. Although this meta‐analysis does not allow us to generate new theories in the open innovation literature, it offers valuable insights to both academics and practitioners regarding factors affecting the implementation of open innovation and its performance implications.