Stomatitis is a troublesome adverse effect of disease-modifying anti-rheumatic drug (DMARD) therapy in rheumatoid arthritis (RA) patients. This review presents data to examine the incidence, clinical features and consequences of DMARD-related stomatitis, and suggests an algorithm for its clinical management. The specific objectives of the two studies presented here were to determine the incidence of DMARD-related stomatitis and its effect on DMARD continuation, and secondly to identify the clinical and laboratory risk factors. We investigated two cohorts of patients: (i) a retrospective survey of data collected from drug monitoring clinics run for patients on DMARDs from 1987 to 1994 involving 1539 patients and 2394 drug exposures; (ii) a prospective study of 25 consecutive RA patients presenting with DMARD-related stomatitis compared to 29 RA controls with no history of DMARD stomatitis. The retrospective survey showed that 2% of DMARD patients stopped therapy because of stomatitis, but 55% of these were able to resume the same therapy. In the case control study. 24% of patients discontinued temporarily and 8% permanently. Cases of DMARD-related stomatitis differed from controls in that they had a higher incidence of previous mouth ulcers (40% vs 14%), they smoked less (8% vs 31%) and Schirmer's test was more often abnormal (44% vs 21%). There were no differences in RA severity, disease activity or oral hygiene. Haematinic deficiencies were equally common in cases and controls: 30% for iron, 8% for vitamin B12 and 24% for folic acid. Herpes simplex virus was involved in a minority (8%) of cases. In conclusion, the occurrence of stomatitis in RA patients on DMARD should not lead to cessation of drug therapy, but to a careful evaluation so that patients may be maintained on effective treatment.