The article outlines a history of the concept of "disease-modifying antirheumatic drugs" or DMARDs--from the emergence in the 1970s of the idea of drugs with decisive long-term effects on bone erosion in rheumatoid arthritis (RA), through the consolidation and popularisation in the term DMARD in 1980s and 1990s. It then examines the usage of the terms "remission-inducing drugs" (RIDs) and "slow-acting anti-rheumatic drugs" (SAARDs), which for some years offered competition to the term DMARDs, thus underscoring the contingency of the establishment of DMARD as a word. Finally, it juxtaposes the apparently spontaneous emergence of the three terms DMARD, SAARD and RID, and the disappearance of the latter two, with a failed attempt in the early 1990s to replace these terms with the new term "disease-controlling antirheumatic treatment" (DC-ART). The analysis highlights the paradoxical qualities of the DMARD concept as robust albeit tension ridden, while playing down the role of identified individuals and overarching explanations of purpose.