Brain atrophy is recognized as a key component of multiple sclerosis (MS) but its reliable measurement at the individual level remains challenging. Besides physiological variability, differences between MRI scanners and acquisition methods are particular hurdles to reliably measure within-patient brain volume changes longitudinally. There is inconsistency between atrophy measurements of a patient scanned using different MR systems or at the same system but using different measurement (i.e., image acquisition) methods. This inconsistency decreases the statistical power in multi-center studies and hampers the use of atrophy measurements in personalized diagnosis, prognosis and treatment of the MS patients. This dataset contained 30 pwMS and 10 HC to validate a new standardization method to reliably measure atrophy across scanners and acquisition methods. After this validation, data from MS patients across the world can be accumulated and combined into a reference database that can be used. Participants had a scan and rescan on three different MR scanners