OBJECTIVES: To compare the cost-effectiveness of current disease modifying therapies (DMT) for patients with relapsing-remitting multiple sclerosis (RRMS) in the US. METHODS: An economic model was created to predict the course of patients with RRMS following initiation of a DMT.  Natalizumab (NAT), dimethyl fumarate (DMF), and peginterferon beta-1a (PEG), were compared with fingolimod (FIN), glatiramer acetate (GA, 40 milligrams thrice weekly), and interferon beta-1a (INT, 44 micrograms thrice weekly), respectively. The Markov state transition cohort model predicted disease progression across RRMS Expanded Disability Status Scale [EDSS] states and for secondary-progressive (SPMS) EDSS states in 3-month cycles over a 10-year time horizon. The patient cohort was at risk of death, relapse, or discontinuation (due to reaching EDSS level 7, or following DMT-specific rates) in each cycle. Outcome measures were relapses, relapse-free time, MS progression, and progression and clinical disease activity-free years. Costs included drug, administration, monitoring, relapse, and EDSS state costs. Incremental cost-effectiveness ratios (ICERs) were estimated for each of the outcome measures. RESULTS: Costs ranged from $477,158 (DMF) to $526,667 (INT). NAT, DMF, and PEG were less expensive with equal, or better, outcomes.  NAT and DMF were dominant (less costly and more effective) compared to FIN and GA, respectively, for all ICERs. PEG dominated INT on progression and clinical disease activity outcomes. Comparable relapse-related outcomes cost more than $30,000 with INT compared to PEG. Variability in drug costs and parameters that affected drug cost accrual (eg, discontinuation rates and the decision to drop out after SPMS conversion) had a considerable impact on ICERs. CONCLUSIONS: Results from this analysis suggest that the NAT, DMF and PEG are cost-effective DMT choices compared to FIN, GA, and INT, respectively. The actual impact to a particular plan will vary based on drug pricing and other factors affecting drug cost accrual.