In this paper 4 studies on the course of delusional disorders of midlife, partly-diagnosed as ‘late schizophrenias’ in the sense of M. Bleuler, published up to the 70s and 3 studies on the same topic which were published in the 80s, all European, are analysed with regard to differences in their results in order to find out factors operating on these differences and to propose standards to be followed in such studies. Four such factors are stressed: different follow-up times leading to different rates of loss of probands by death, unreliable use of diagnostic terms and, probably more influential, unreliable use of terms describing the course, and different preconditions of the recruiting process by differences in the care systems the probands are collected in. Four conclusions are derived from this finding: the need for comprehensive and valid information, for reliable use of terms describing the course, for information about the care system the patients come from as an important precondition of the process of recruiting probands, and probably for the further development of a ‘language’ for the description of