OBJECTIVES: Disease-modifying anti-rheumatic drugs (DMARDs) are the cornerstone of rheumatoid arthritis (RA) pharmacotherapy and should be initiated promptly after RA diagnosis. We examined trends in DMARD use among RA patients in Quebec, and factors correlated with DMARD initiation in newly diagnosed RA. METHODS: Quebec administrative health databases were used to identify RA subjects and their claims for medical and pharmaceutical services between January 1, 2002 and December 31, 2008. To describe DMARD use, cross-sectional analyses were performed on November 1 of each year. For subjects newly diagnosed with RA, multivariable logistic regressions were used to identify possible predictors of DMARD initiation at 12 months and Kaplan-Meier curves to define the probability of initiating a DMARD over time. RESULTS: A total of 32,533 subjects were included (mean age: 67.5 years; 70.4% female). Over the study period, the percentage of subjects on a DMARD increased from 42.0% (November 2002) to 43.2% (November 2008). Being followed by a rheumatologist (vs. GP) was the strongest predictor of DMARD initiation (OR=4.39; 95%CI: 3.80-5.08). The use of NSAIDs, corticosteroids, and opioids in the year prior to cohort entry and the calendar year of cohort entry had a positive effect on DMARD initiation, whereas age, comorbidity score, and the use of acetaminophen had a negative effect. For biologics, calendar year was the strongest predictor (OR 2007 vs. 2002=10.78; 95%CI: 2.45-47.37). Of subjects newly diagnosed in 2002, 0.1% had a biologic initiated within one year, while for those newly diagnosed in 2007 the percentage was 1.3%. In any newly diagnosed subjects, averaged over 2002-2007, the probability of having initiated any DMARDs at 12 months was 38.5% (47.8% for those followed by a rheumatologist). CONCLUSIONS: Despite encouraging signs for earlier aggressive RA management, DMARD use appears to be sub-optimal in Quebec. Use of DMARDs was much higher among subjects followed by a rheumatologist.