Background : The Minimal Assessment of Cognitive Function in Multiple sclerosis (MACFIMS) is an internationally recognised battery of neuropsychological tests for patients with multiple sclerosis (MS). Objectives : To establish regression-based norms for the MACFIMS in French-speaking healthy subjects (HS) and validate its use in persons with multiple sclerosis (PwMS). Methods : 136 PwMS, including 43 with relapsing-remitting MS, 46 with secondary progressive MS and 45 with primary progressive MS, as well as 276 HS were enrolled. Regression-based norms and validity were established for the seven tests of the MACIMS: the Symbol Digit Modalities Test (SDMT), the Paced Auditory Serial Addition Test (PASAT), the French learning test (FLT) a French-adapted memory test (or the California Verbal Learning Test (CVLT) at re-testing), the Judgment of Line Orientation Test (JLO), the ‘épreuve de classement de cartes de Champagne’ (ECCC), a French adaptation of the DKEF-sorting test, the Brief Visuospatial Memory Test (BVMT-R) and the Controlled Oral Word Association Test (COWAT). Results : Regression-based norms of MACFIMS tests were established in the HS population. The MACFIMS battery was able to identify cognitive impairment (CI) (at least two abnormal tests in different domains) in 32.7% of PwMS. The domains with more frequent impairment were (in descending order): learning followed by IPS, delayed memory, verbal fluency and working memory. Conclusion : This study established the regression-based norms for French subjects of the French adaptation of the MACFIMS and its validity in PwMS.< Réduire