Abstract: This study investigates how academic-industry collaboration experiences influence scientists' generation of outlier innovations. By analyzing patent data from 1993 to 2018 in China and employing a panel negative binomial random effects regression model, this research examines the impact of academic-industry collaboration on scientists' production of outlier innovations and its boundary conditions. The results indicate that academic-industry collaboration experiences significantly enhance the likelihood of scientists discovering outlier innovations, possibly because such collaboration improves scientists' understanding of industry goals and thinking patterns while also stimulating their attention to the commercialization of research outcomes. Furthermore, scientists' organizational search resources, such as individual innovation experience and collaboration networks, moderate the effect of academic-industry collaboration experiences on the possibility of outlier innovation’s origination. Specifically, the wider the scientists' breadth of innovation experience, the weaker the positive impact of academic-industry collaboration experiences; similarly, the larger the number of indirect collaborators within the organization, the weaker the positive influence of academic-industry collaboration. This study provides a new perspective for understanding the role of scientists in the innovation ecosystem and offers managerial implications for facilitating the transformation of scientific research into technological innovation.