The psychometric properties of the Vineland Adaptive Behavior Scales Survey Form were studied in a total population of children and adolescents with MR, and in the specific levels of functioning (n=826, age 4-18 years). The original division into (sub)domains, as assigned by the authors, was replicated in the total population and in the mild and moderate levels of functioning. In the severe and profound levels of functioning the structure was less well recognized. The reliability of the instrument proved to be good in the total population and the subgroups. The construct validity was high in all groups. The implications of these findings are discussed with respect to the usefulness of the Vineland for the population with MR.