This study integrates text mining into discourse analysis to provide a corpus-assisted discourse investigation of crisis communication during the pandemic in the CSR reports of leading corporations in China and the US. Drawing on the theories and methods of text mining, corpus linguistics, and systemic functional linguistics, this research reveals not only the foregrounding of danger as a prominent theme in Chinese and American CSR reports but also their distinct preferences in representing these dangers. The findings highlight not only cultural differences in crisis communication during the pandemic but, more importantly, the potential for a balanced integration, or “synergy”, of systemic functional linguistics and text mining in corpus-assisted discourse studies.