Background: Tofacitinib (TOFA) was the first Janus kinase inhibitor (JAKi) to be approved for the treatment of rheumatoid arthritis (RA). However, data on the retention rate of TOFA therapy are still far from definitive. Objective: The goal of this study is to add new real-world data on the TOFA retention rate in a cohort of RA patients followed for a long period of time. Methods: A multicenter retrospective study of RA subjects treated with TOFA as monotherapy or in combination with conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) was conducted in 23 Italian tertiary rheumatology centers. The study considered a treatment period of up to 48 months for all included patients. The TOFA retention rate was assessed with the Kaplan–Meier method. Hazard ratios (HRs) for TOFA discontinuation were obtained using Cox regression analysis. Results: We enrolled a total of 213 patients. Data analysis revealed that the TOFA retention rate was 86.5% (95% CI: 81.8–91.5%) at month 12, 78.8% (95% CI: 78.8–85.2%) at month 24, 63.8% (95% CI: 55.1–73.8%) at month 36, and 59.9% (95% CI: 55.1–73.8%) at month 48 after starting treatment. None of the factors analyzed, including the number of previous treatments received, disease activity or duration, presence of rheumatoid factor and/or anti-citrullinated protein antibody, and presence of comorbidities, were predictive of the TOFA retention rate. Safety data were comparable to those reported in the registration studies. Conclusions: TOFA demonstrated a long retention rate in RA in a real-world setting. This result, together with the safety data obtained, underscores that TOFA is a viable alternative for patients who have failed treatment with csDMARD and/or biologic DMARDs (bDMARDs). Further large, long-term observational studies are urgently needed to confirm these results.