Only some of the areas of drug interactions of relevance to those treating rheumatic diseases have been mentioned and there are still enormous gaps in our knowledge. It is likely that some potential areas of danger have been over-emphasised, being based on speculation rather than real data or purely animal experiments using non-clinical doses of drugs. We are learning how certain drugs can stimulate or inhibit the metabolism of other drugs through effects on liver enzymes systems. For example, the metabolism of corticosteroids is induced by phenylbutazone (and also by phenobarbital and phenytoin). The patient with active rheumatoid arthritis with a low serum albumin would be unusually susceptible to changes induced by combinations of highly protein-bound anti-inflammatory drugs. Finally, although drug interactions are responsible for adverse effects it has been suggested that a more frequent cause of therapeutic failure is not drug interactions but the increase in the number of drug defaulters when more than one drug is prescribed.