Objective: In patients suffering multiple sclerosis activity despite treatment with interferon β or glatiramer acetate, clinicians often switch therapy to either natalizumab or fingolimod. However, no studies have directly compared the outcomes of switching to either of these agents. Methods: Using MSBase, a large international, observational, prospectively acquired cohort study, we identified patients with relapsing-remitting multiple sclerosis experiencing relapses or disability progression within the 6 months immediately preceding switch to either natalizumab or fingolimod. Quasi-randomization with propensity score-based matching was used to select subpopulations with comparable baseline characteristics. Relapse and disability outcomes were compared in paired, pairwise-censored analyses. Results: Of the 792 included patients, 578 patients were matched (natalizumab, n = 407; fingolimod, n = 171). Mean on-study follow-up was 12 months. The annualized relapse rates decreased from 1.5 to 0.2 on natalizumab and from 1.3 to 0.4 on fingolimod, with 50% relative postswitch difference in relapse hazard (p = 0.002). A 2.8 times higher rate of sustained disability regression was observed after the switch to natalizumab in comparison to fingolimod (p<0.001). No difference in the rate of sustained disability progression events was observed between the groups. The change in overall disability burden (quantified as area under the disability-time curve) differed between natalizumab and fingolimod (20.12 vs 0.04 per year, respectively, p<0.001). Interpretation: This study suggests that in active multiple sclerosis during treatment with injectable disease-modifying therapies, switching to natalizumab is more effective than switching to fingolimod in reducing relapse rate and short-term disability burden.