ObjectiveMethods of cognitive measurements in multiple sclerosis (MS) are not standardized. We aimed to identify the prevalence of cognitive domain‐specific impairment (DSI) in MS by using subtests of the Brief Repeatable Battery of Neuropsychological Tests (BRB‐N) with analyzing different cutoff values.MethodsThe systematic review and meta‐analysis were registered on PROSPERO (ID: CRD42021287004). The systematic literature search was performed via PubMed, Embase, and CENTRAL on 24 October 2021. Inclusion criteria were adults of different MS subtypes (CIS, RRMS, PPMS, and SPMS) with the condition of distinct DSI measured by BRB‐N. Pediatric MS, computerized versions of BRB‐N, and patients receiving steroids were excluded. Primary outcome was pooled prevalence rates of impaired patients within each cutoff and MS subtype, with 95% confidence interval, I‐squared statistics for heterogeneity, and chi‐squared test for subgroup differences. Risk of bias was assessed using the “JBI Quality Assessment Tool for Prevalence Studies.”ResultsIn 48 eligible observational studies (n = 3431 patients), the three most prevalent thresholds were the 2.0 SD and 1.5 SD below the mean of normative values, and the score below the fifth percentile of the normative values. A progressively increasing worsening of the overall DSI was observed from CIS, moving toward RRMS, PPMS, and SPMS.InterpretationCognitive impairment is observed in all MS phenotypes, with varying degrees. Due to several potential influencing factors, our comprehensive literature review has not revealed consistent findings, and we, therefore, recommend considering a more sophisticated, “individual referencing” approach, acknowledging the diverse clinical and sociodemographic characteristics among populations and disparities in cognitive testing.