Abstract: Exploring the effect of basic research on firms’ scientific breakthroughs can help firms asses the effectiveness of investments in basic research. However, the lack of fine-grained measurement method of basicness has hindered this research. To this end, using Level Score (LS) to measure the basicness of a single paper, this paper explores the influence of basic research on firms’ scientific breakthroughs using regression analysis based on the publication data of global biomedical firms from 2015 to 2017. The results show that, first, the number of basic research papers and applied research papers referenced by biomedical firms can significantly and positively affect the number of papers they produce, in line with the basic input-output relationship. Second, basic knowledge plays a key role in the output of firms’ scientific output. Enhancing the average basicness of knowledge inputs facilitates biomedical firms to produce more scientific breakthrough papers. Finally, we also find that in terms of research types, a more diverse knowledge combination increases the likelihood of producing scientific breakthroughs. This paper empirically validates the contribution of basic research to scientific innovations. In addition, this paper can provide a scientific basis for firms in the field of biomedicine to increase their investment in basic research and enhance their basic research capacity, and also provide certain insights for the formulation of related scientific and technology policies.