OBJECTIVE：To provide reference for pharmaceutical enterprises to better fulfill their social responsibilities . METHODS：Taking the pharmaceutical enterprises listed in Shanghai and Shenzhen stock markets as the research objects ，the social responsibility texts and environmental protection texts of the included enterprises from 2017 to 2019 were classified and analyzed according to the three dimensions of time ，ownership and drug production type by using ROST Content Mining 6.0 software. The semantic network and node centrality were analyzed. RESULTS ：From the social responsibility texts of the included enterprises，it could be seen that from the perspective of time dimension ，the included enterprises paid more attention to “employees”，“social responsibility ”and“development”in 2017-2019，and the high-frequency words in the three years were relatively stable. From the perspective of ownership ，state-owned holding enterprises paid more attention to “development”and “targeted poverty reduction ”，while private holding enterprises paid more attention to “employees”，and collective holding enterprises paid more attention to “social responsibility ”. From the perspective of drug production type ，the word frequency of “targeted poverty reduction ”of TCM pharmaceutical enterprises was higher. From the environmental protection texts of the included enterprises，it could be seen that from the time dimension ，the included enterprises paid more and more attention to environmental protection in recent years ，and the description of environmental protection in the texts was more detailed. From the perspective of ownership，private holding enterprises paid more attention to environmental protection than collective holding enterprises and state-owned enterprises. From the perspective of drug production type ，due to the different types of drugs ，chemical drugs ，TCM， biological drugs and active pharmaceutical ingredients enterprises had their own characteristics in pollutants and their treatment processes. The results of semantic network and node centrality analysis showed that ，“development”，“protection” and “targeted poverty reduction ”were three high-frequency words with high centrality in the semantic analysis network of social responsibility text ，the middle centrality was 5.555，1.695，1.521 respectively；“emission”and“testing”were two key words with high centrality in the semantic analysis network of environmental protection texts ，the middle centrality was 3.856， 1.914 respectively，which further reflected that the included enterprises paid more attention to the above issues in terms of social responsibility performance and environmental protection. CONCLUSIONS ：The listed pharmaceutical enterprises have a broad consensus on social responsibility in China ，and have well fulfilled their social responsibility ，which is well combined with the current social hot spots. For environmental protection ，as a high pollution industry ，the pharmaceutical enterprises also performs the responsibility of environmental protection.