Disability worsening can occur at variable rates for patients with multiple sclerosis (MS), so machine learning represents a potential tool for predicting disease course. Tanuja Chitnis, MD, Brigham and Women’s Hospital, Boston, MA, discusses a study of over 1000 patients with MS from the CLIMB study and the EPIC dataset, to assess the predictive features of various machine learning algorithms and ensemble learning approaches. The models were assessed on their ability to predict expanded disability status scale (EDSS) worsening or non-worsening at up to 5 years post-baseline, and ensemble learning methods were found to be more effective and robust predictors compared to the standalone machine learning algorithms. These techniques increased accuracy for disease course prediction, achieving 0.79 and 0.83 AUC scores for the CLIMB and EPIC datasets, respectively. This interview took place during the ACTRIMS Forum 2021.