Abstract: In the era of knowledge economy, universities have become an important source to provide key technical support for the industry to cope with global competition, via such mechanisms as patent licensing agreement. With the increasing patent behavior, universities are facing more pressure from multi capital and commercialization. This phenomenon has aroused extensive discussion among scholars. How do universities balance the public attribute of academic research and the private attribute of patent commercialization? Does university patent behavior hinder the free flow of scientific knowledge and reduce the academic influence of researchers? In the context of the rapid growth of patent behavior in China’s universities, it is necessary to deeply explore the relationship between patent behavior and academic influence to find the balance between commercialization of technology and publicity of academic research, which has important theoretical and practical significance to improve the science and technology evaluation system and innovation efficiency in China. Large numbers of vital empirical studies have been exploring the relationship between academic research and patent commercialization. Nevertheless, the debate over whether patent behavior helps or restricts university researcher’s academic influence has always been an argument. On one hand, some believe that researchers involved in patent behavior can benefit both from the networks of patent and academic, such as research projects are more related to the industry and publications are more likely to be accepted by high quality journals. On another hand, some argue that patent behavior may change researchers’ interest by potential commercial value other than academic value, which leads negative effect on the future of scientific research. The existing literatures mainly focus on the stage of patent application, however, less attention has been paid on the effect of patent licensing stage. As we know, the stage of patent is not static, but changes with time. Therefore, there is a great necessity to further study how the whole procedure of patent behavior influence university researchers’ academic influence. In order to better study the impact of university researchers' patent behavior on their academic influence, firstly, this study divides patenting into two stages according to China's patent system: "application-granting" stage and "granting-licensing" stage, which expands the previous research perspective from one patent stage to the whole procedure of patent behavior. Second, we establish a database collecting the data of patent behavior and academic influence of multi-disciplinary researchers in 112 “211” universities and provincial and ministerial-level universities from 2008 to 2017, which includes individually micro data and provides fundamental support for relevant study in China’s background. Third, this paper comprehensively uses negative binomial regression and binary logistic regression to empirically analyze the influence of university researchers' patent behavior on their academic influence of "application-granting" stage and "granting-licensing" stage, which reveals nonlinear relationship between patent behavior and academic influence from multi dimension. Finally, corresponding optimization suggestions are proposed according to the research results． The results show that: (1) In "application-granting" stage, there is a significant inverted U-shaped relationship between the number of patent applications and academic influence. With the increase of patent applications, the positive impact on academic influence will increase. Nonetheless, there will be a negative impact on scholars’ citation, H index and G index when the number of patent applications reaches about 35 per year. As for the academic title, the turning point of patent applications is around 38 per year. (2) In "granting-licensing" stage, when university researchers use exclusive license to trade patents, there will be a significant inhibitory effect on academic influence, no matter measuring citation, H index, G index or high-level academic title. The probability of university researchers obtaining high-level academic title is reduced by 70.4% if researchers’ patent is licensed exclusively. Citation is the most affected by exclusive licensing, followed by G index, and H index is the least affected. These findings provide reference for the whole process management of university patents and the formulation of relevant policies. Corresponding optimization suggestions are proposed according to the micro quantitative evidence provided by this study. (1) University researchers should be encouraged to participate in patent behavior properly, which can fully use the cooperative effect and avoid the crowding out effect caused by too many patent applications. (2) Exclusive license should be considered carefully when university researchers trade their patents, especially in the condition that the patents are supported by public resources such as government funding. (3) The government should provide investment and financing support for the commercialization of university patents to promote and protect original innovation. This study is of great theoretical and practical significance to clarify the mechanism of how patent behavior effects university researchers’ academic influence, to answer the question of how universities could achieve breakthroughs in original innovation and provide innovation force for China's economic development at the same time, and to optimize the construction of China’s innovation system.