Abstract: Enterprise basic research on enterprises is an important cornerstone of the construction of a technological country, which helps to promote the deep integration of scientific research and industrial development. Compared with developed countries, there exist a lack of basic research investment in Chinese enterprise. To improve enterprise’s investment in basic research is vital for indigenous innovation in China. Based on the perspective of knowledge spillover, this study constructs the correlation data between science publication and patent science citations of Chinese manufacturing listed companies from 2000 to 2022, exploring the knowledge spillover mechanism of enterprise technology literature on focal enterprise patents and on competitive enterprise patents. We suggest that basic research can be viewed as the input of applied research and development. However, knowledge spillover may result in incentive distortion of enterprise’s basic research investment. By distinguishing knowledge internal use and knowledge spill-out, we study the incentives, mechanisms and market effect of Chinese enterprise’s basic research. The sample of the research is listed manufacturing enterprises in China’s A share stock market. We use the enterprises which has at least one year of positive R&D investment and at least one patent application. This gives us a non-balanced panel of 29247 firm-year observations with 2902 unique firms. Patent data is from Incopat database. Publication data is from CNKI database. Financial data is from CSMAR database. We match data from different databases. To identify the relationship between patent citations in enterprise basic research publications, first, 186148 patents with scientific citations were obtained after cleaning the original 1.8 million patents. Considering the situation where multiple patents refer to the same scientific literature, the original information of cited scientific literature was ultimately cleaned out. Secondly, considering that there may be some errors in the titles of scientific papers in CNKI and Incopat databases, fuzzy matching was performed on the title of articles and patent cited scientific literature in CNKI. Using Natural Language Processing (NPL) and Python_ Gensim library, we calculate the similarity between two database paper name strings. Using jieba word segmentation on the literature title strings of the two datasets, we construct an index dictionary of corpus features, and use a word bag model to construct a similarity matrix for fuzzy matching. We use twoway fixed effect model in our empirical research. Research findings are: (1) Knowledge internal use promotes focal enterprises to conduct basic research, while knowledge spill-out decreases focal enterprises from conducting basic research. (2) The citation of competitive enterprise patents to focal enterprise science literature can help improve the innovation performance of focal enterprises. (3) Compared with knowledge internal use, knowledge spill-out in basic research significantly decrease the financial performance of the focus enterprises. Robustness checks using Heckman model gain the same results. Heterogeneity test indicates that knowledge spillovers has a more significant impact on the basic research of non-state-owned enterprises. Basic research knowledge internal use promotes the applied research of high-tech industries and small-scale enterprises, while basic research knowledge spill-out promotes the applied research of non-state-owned enterprises and large-scale enterprises. This article reveals the incentives, spillover mechanisms, and market effects of Chinese enterprise basic research. The conclusions provide policy implications for consolidating the important position of enterprises in scientific and technological innovation, and guiding the deep integration of basic research and applied research.