Despite prolonged survival, patients with multiple sclerosis (MS) experience considerable morbidity, which adversely impacts quality of life. To assess the risk-benefit of a clinical trial of high dose immunosuppressive therapy with autologous hematopoietic stem cell transplantation for MS, we sought to determine the natural history of the disease in a comparison group of untreated patients. We identified 285 individuals with 2132 combined observation years (median: 5.6 years; 5th to 95th percentile: 1-21 years), with Expanded Disability Status Scale (EDSS) scores of 3.0-5.5 at baseline observation. Disease-related mortality was zero at five years, 5.4% at 10 years, and 22% at 15 years (40 patients contributing to the data point; 95% confidence interval: 4-32%). Risk for progression to advanced disability, defined as an EDSS score of 8, was very low for the subgroup with a baseline EDSS score of 3-3.5; however, for those with a baseline EDSS score of 4-5.5, 3% had advanced disability after two years, 5% after three years, 6% after four years, 12% after five years, and 40% after 10 years. The estimated probability of disease progression, defined as an increase in EDSS score by > or = 1.0 sustained for at least 180 days, was 5% after one year, 14% after two years, 22% after three years, 38% after five years, 57% after 10 years, and >80% after 20 years of observation. The relevance of these features to the design of the clinical trial is discussed.