PURPOSE: To evaluate the attribution of false belief in individuals with Down syndrome. METHODS: Eleven children of both genders with Down syndrome and ages between four and eight years composed the Down Group (DG). All subjects used verbal communication, had mild to severe mental retardation, and were attended at the same institution. In addition, 85 children within normal development with ages between four and six years were recruited at an elementary school, constituting the Control Group (CG). The Peabody Picture Vocabulary Test (PPVT) was used to compare the level of verbal comprehension of the groups, and the adaptation of the "smarties test" was used to assess the attribution of false belief. RESULTS: The analysis of the PPVT showed a statistically significant difference between the groups, with the DG presenting scores below the third standard deviation, and the CG, below the first standard deviation. Regarding the false belief analysis, the CG presented increasing scores in all questions as the age group increased. The same was not observed for the DG, whose subjects with best results were those that attended speech-language therapy for a longer time at the institution. No correlation was found between receptive vocabulary level and the ability of false belief. CONCLUSION: The CG showed a better performance in all questions of the false belief test, when compared to the DG. Thus, it was possible to evaluate false belief in children with Down syndrome.