Background: Fingolimod is an efficacious treatment for relapsing-remitting multiple sclerosis (RRMS) and there is class I evidence that it is superior to standard care in reducing relapse rate. However, real-world data investigating its effectiveness and potential predictors of response are still scarce. Objective: To estimate (i) the proportion of fingolimod-treated patients who achieved the no evidence of disease activity (NEDA-3) status; and (ii) to determine which baseline (i.e. at treatment start) clinical and magnetic resonance imaging (MRI) variables were associated with better outcomes. Methods: We collected clinical and MRI data of RRMS patients treated with fingolimod and followed-up for 24 months. The proportion of patients who had NEDA-3 -i.e. absence of relapses, sustained Expanded Disability Status Scale (EDSS) worsening and radiological activity on MRI -was estimated. A Cox proportional hazard model was carried out to investigate which baseline characteristics were associated with the NEDA status at follow-up. Results: We collected data of 201 patients who started fingolimod. Of them, 24 (12%) were treatment-naïve, 115 (58%) were switched after failing a self-injectable drug, and 60 (30%) switching from natalizumab. Five patients who discontinued fingolimod early (within 3 months) (bradycardia, n = 2; leukopaenia, n = 2; macular oedema, n = 1) were removed from the analysis. At follow-up, 118 (60%) patients achieved the NEDA-3 status, while 78 experienced clinical and/or MRI activity. The risk of not achieving the NEDA-3 status was associated with higher baseline EDSS score (hazard ratio [HR] = 1.18, p = 0.024) and more relapses in the 12 months prior to fingolimod start (HR = 1.61, p = 0.014). Conclusion: Our findings suggest that fingolimod may lead to a better control of the disease if started in patients with a less aggressive disease (i.e. fewer pre-treatment relapses and milder disability level), thus supporting its possible role as an early treatment for MS.