OBJECTIVES:  This study compared different algorithms to identify patients with multiple sclerosis (MS) in claims data and recommended the most appropriate algorithm. METHODS: Our literature review on MS studies in claims data identified ten different algorithms to identify MS patients. Some algorithms require either MS diagnosis or MS treatment, or both; some require two or more diagnoses or treatment claims; some require evidence of other neurological conditions in addition to an MS diagnosis. These algorithms were used to identify MS patients in Truven Health MarketScan®Commercial and Medicare Supplemental Databases in 2004-2011. For each algorithm, MS prevalence rate, patients’ age and gender, proportion of patients with magnetic resonance imaging (MRI) and MS treatment were examined and compared with those in published studies wherever possible. RESULTS:  Two algorithms identified about the same number of MS patients: one algorithm required ≥2 MS diagnoses ≥30 days apart (123,064 patients were identified) and the other required ≥1 principal inpatient MS diagnosis or ≥2 MS diagnoses ≥30 days apart (123,160 patients were identified). Both populations had a mean age of 47 and 76% female, consistent with that reported in the literature (mean age: 40.9-50.3, female: 66-80%); a total of 69% of them had MS treatment. The proportion of patients with MRI increased from 41% in January 1, 2004 - February 28, 2005 to 66% in March 1, 2005 – September 30, 2011, consistent with when McDonald’s criteria and update were published. These two algorithms also produced a prevalence rate of 135 per 100,000 people, same as the rate reported by Atlas of MS Database and the National MS Society. Prevalence rates based on other algorithms were either too high or too low. Thus these two algorithms were the most appropriate to identify MS patients in claims data. CONCLUSIONS:  Comparison of available patient and epidemiological characteristics with published literature suggests that MS patients can be accurately identified in claims data.