We still know relatively little about how researchers use qualitative data analysis software (QDAS) such as ATLAS.ti and NVivo. We conducted a discourse analysis of 763 empirical articles published from 1994 to 2013 that explored the language used by researchers when reporting QDAS use. We found that most researchers provided few details of their QDAS use beyond naming the program, but the detailed accounts provided by some authors provided valuable insights into the ways that QDAS programs can be used to support data analysis and the reporting of research outcomes. We conclude with suggestions for best practices in reporting QDAS use. We encourage researchers to provide more detail about their program usage, e.g. by choosing active rather than passive voice to avoid attributing agency to the software, defining specialized QDAS terminology to prevent confusion, and avoiding unsubstantiated claims of a relationship between QDAS use and improved quality.