Preschoolers (n = 20), second graders (n = 20), fifth graders (n = 20), and graduate-level college students (n = 20) were asked to determine the reality status of specific items, such as dolls and dreams, and to generate examples of real and not-real things. The criteria used for determining whether or not specific items were real differed significantly across the age range sampled and were strongly influenced by the type of items being judged. Criteria clearly changed with age, with item-specific criteria being substituted for more general verification criteria.