OBJECTIVES: Several new agents have shown positive results in clinical trials for prevention of relapses in multiple sclerosis. The objective of this study was to compare network meta-analysis results with traditional meta-analysis. METHODS: A systematic literature search for randomized clinical trials for multiple sclerosis was undertaken for the databases Pubmed, Embase, Biosis, Google Scholar and Cochrane. Data was collected for the study type, methods, country and key findings. Extracted study data included study design, patient charatcertristcis and MS related outcomes. A bayesian random effects network meta-analysis (NMA) model was developed in WinBUGS14. Results were compared to previously published traditional meta-analysis (TMA). RESULTS: We identified 730 references and found 10 randomized trials in 4095 patients with 2518 events. The drugs included in our study were interferon (IFN), glatiramer (GLAT, natalizumab (NAT) and fingolimod (FING). The Odds Ratios (ORs) for FING versus NAT was 1.66 (0.31 – 9.96),FING versus GLAT 2.19 (0.50 – 7.49), FING versus IFN 2.25 (0.90 – 5.57), FING versus PLB 4.42 (1.46 – 14.87), NAT versus GLAT 1.32 (0.22 – 5.40), NAT versus IFN 1.321.35 (0.29 – 5.75), NAT versus PLB 2.68 (0.71 – 9.99), GLAT versus IFN 1.03 (0.42 – 3.16), GLAT versus PLB 2.04 (0.95 – 5.91), IFN-B versus PLB 1.97 (1.01 – 4.14). The probability to be best in preventing relapses was 75.9% for FING, 17.0% for NAT and 6.3% for GLAT. Compared to TMA, the effect sizes were significantly larger in NMA. CONCLUSIONS: Network meta-analysis of multiple sclerosis trials shows fingolimod as the best treatment. Compared to traditional meta-analysis, network meta-analysis enables more robust estimation of effect sizes.