Abstract Background This study aimed to compare the effects of teriflunomide and ocrelizumab on clinical and MRI endpoints related to smoldering activity in relapsing–remitting multiple sclerosis (RRMS). Methods In this observational, longitudinal, multicenter study, we included 128 people with RRMS (pwRRMS) treated with teriflunomide and 495 treated with ocrelizumab. Outcomes included time to progression independent of relapse activity (PIRA). In a subset, we also assessed brain volume loss (BVL), longitudinal changes in diffusion tensor imaging (DTI) metrics, and the burden of paramagnetic rim lesions (PRLs). Propensity score matching was used for between-group comparisons. Results Over a median follow-up of 3.1 years in the ocrelizumab group and 1.9 years in the teriflunomide group, there were no significant differences in the risk of PIRA (HR for teriflunomide vs. ocrelizumab: 0.80 [95%-CI:0.40–1.60]; p = 0.53). PwRRMS treated with teriflunomide exhibited lower annualized rates of BVL (−0.80 [95%-CI: −0.91; −0.69] vs. −1.06 [95%-CI: −1.25; −0.86]; p = 0.025) and gray matter volume loss (−0.92 [95%-CI: −1.05; −0.79] vs. −1.20 [95%-CI: −1.43; −0.97]; p = 0.035). No differences were observed in DTI metrics or PRL count. Conclusions This real-world study suggests that teriflunomide shows similar efficacy to ocrelizumab on smoldering activity, with a potentially greater effect in reducing BVL. Further research is needed to confirm these findings and understand their long-term implications.