Purpose When a comprehensive neuropsychological assessment cannot be carried out, a quick and discriminant tool of good psychometric properties can be useful to practitioners. The purpose of this paper is to examine the use of the Montreal Cognitive Assessment (MoCA) in patients with Prader–Willi syndrome (PWS) and to test its reliability for cognitive assessment in a population with intellectual disabilities. Design/methodology/approach Thirty-seven adults with PWS took the MoCA. Reliability of the battery was tested using Cronbach’s alphas. The performance of PWS adults in each subtest was then compared to that of a normative population of healthy adults. Findings The MoCA was found to be unreliable in PWS. The subtests analyses indicated that the PWS sample underperformed the normative population of healthy adults on most subtests of the MoCA. A sub-sample aged between 17and 29 years showed normal performance on Naming and Memory, and a sub-sample aged from 30 to 39 years showed similar performance on Language, Memory and Orientation relative to age-matched normative healthy adults. Research limitations/implications Results showed that the current version of the MoCA, if taken as a whole test for cognitive assessment, does not present with adequate psychometric properties, which the authors interpret as reflecting the heterogeneity in PWS cognitive profiles. If used in PWS, the MoCA may however be useful in examining cognitive functions separately using subtest-based comparisons to normative data. Originality/value This research contributes to a better assessment of cognitive profile in PWS and people with learning disabilities by arguing that the use of psychometric tests should depend more on the specificity of the population under evaluation.< Réduire