Conservation judgments are based on a combination of logical necessity and empirical belief. Results of two experiments support the view that the logical aspect of conservation is developmentally stable, while the empirical aspect varies widely across problems and individuals because of its dependence on relevant experience. (Author/RD)