Several aspects relevant for the evaluation of encephalotropic substances are presented. A) Memory improving pharmacological effects found in animal studies are often lacking in human studies. B) Global clinical measures of therapeutic success differenciate well between verum and placebo. In contrast to psychological tests, however, objectivity and reliability of global measures is low. C) Diagnosing dementia requires laboratory screening, psychiatric judgment and psychological testing. Psychological theories about age-related decline of different cognitive abilities and psychiatric DSM-III criteria of dementia are however not congruent. One reason for this discrepancy is that psychological theories seldom consider the very old, where dementia is present in sufficient degree. We were able to demonstrate this in a pilot study with 51 residents of a home for the aged, whose average age was 81 years. The subtest Similarities (WAIS), a measure of conceptual ability usually considered to be stable throughout the aging process, showed a marked decline after age 75 in accordance with the psychiatric criterion of dementia, impairment of abstract thinking.