The discrepancy between current and premorbid ability is a relevant indicator of acquired mental impairment, which itself is closely related to general cerebral dysfunction. The use of tests sensitive to cerebral dysfunction, raises relatively few problems compared with tests being resistant that are used to estimate premorbid mental ability. For premorbid ability, verbal tests assessing knowledge, especially vocabulary, have been shown to be valid. A test, possibly more insensitive to brain dysfunction than the ones usually administered, is the multiple choice vocabulary test (MWT = Mehrfachwahl-Wortschatz-Test). At present only German versions are available. They are presented in some detail because of their advantages. Construction of the MWT is simple, and it can be easily administered in about five minutes. The results correlate fairly well with global IQ in healthy adults (median of r = 0.72 in 22 samples) and are more insensitive to current disturbances than such tests as the WAIS vocabulary test. The limitations of premorbid tests with respect to diagnostic validity are discussed. It is concluded, that studies which do not control premorbid intelligence have to be considered as a "malpractice" and should not be accepted by scientists.