Objective: To assess the time to initiation of biologic disease-modifying anti-rheumatic drugs (bDMARDs) in ESPOIR, the French cohort of patients with rheumatoid arthritis (RA), and factors associated with the timing of bDMARD initiation.Methods: In total, 658 patients with early RA satisfying the 2010 ACR/EULAR criteria were included between 2003 and 2005 and followed annually for 10 years (end of follow up: 2013-2015). The timing of bDMARD introduction and predictors of use were analysed by the Kaplan-Meier method based on Cox proportional-hazard models.Results: Overall, 178 patients (31.0%, 95% confidence interval [27.0-34.7]) initiated a bDMARD during the 10-year follow-up, with a mean delay of 43.6 months. The penetration rate was higher during the first 2 years of follow-up (6% between the first and second year, approximately 3.3% each year between the second and seventh year, and<2.0% after the eighth year). The first-used bDMARD was etanercept for 72 patients and adalimumab for 71. On multivariate analysis, Disease Activity Score in 28 joints, radiologic progression and positivity for anti-citrullinated protein antibodies were significantly associated with rapid initiation of a bDMARD (P<0.0001), whereas older age at first joint pain was inversely associated (P<0.0001).Conclusions: Although access to bDMARDs is widespread in France, less than one third of patients with early RA in the ESPOIR cohort initiated a bDMARD over the 10-year follow-up. Poor prognostic factors for RA were associated with more rapid initiation, as expected.