Physicians are sometimes hesitant to use disease-modifying antirheumatic drugs (DMARDs) in elderly patients with rheumatoid arthritis (RA), as they are deemed too fragile, although there are no sufficient scientific evidence. We aimed to compare DMARD treatment retention in early RA patients from the ESPOIR cohort, according to age upon inclusion. Overall, treatment retention was evaluated as the percentage of patients whose DMARDs were not stopped, with stratification by age group: < 50, 50-64, and > 65 years. Survival curves were measured using the Kaplan-Meier method. Of the entire ESPOIR cohort (n = 813), 7% were > 65 years old. Methotrexate (MTX) was used by 521 patients, and was the sole DMARD for 198 patients. MTX treatment retention appeared better in patients > 65 years old compared to < 50 years old [HR 0.45 (0.25; 0.81); p = 0.008, n = 195/198] with adjustment on sex, smoking, positive anti-cyclic citrullinated peptide antibodies, positive rheumatoid factor, body mass index, changes in DAS28 and corticosteroid treatment. The proportion of patients using etanercept (n = 111), and this drug's retention rate, did not differ according to patient age. The proportion of patients treated with adalimumab (n = 104) was significantly higher in patients < 50 years old (p = 0.003), and treatment retention was marginally better among younger patients [HR 1.68 (0.88; 3.22), p = 0.12]. Within the ESPOIR cohort, DMARD retention did not appear to differ according to age-except for better retention of MTX treatment in patients 50-64 years old, and of adalimumab in patients < 50 years old.