SUMMARYDuodenal ulcer is a rather common disease, showing a clear family concentration. The aim of this work was to detect, in the etiology of this disease, possible genetic components, and to study their characteristics. The survey has been carried out by analyzing the data collected in 100 families, assembled by starting from a corresponding number of « index cases », and utilizing for comparison and also further analyzing the data by Doll and Buch (1950) and by Doll and Kelloch (1951). The " secondary cases " found, as a whole, were 109 (90 males and 19 females). The results may be summarized as follows:1) the frequency of the disease in the sibships, independently from sex and corrected for age, is of 32.5% ± 2.5;2) the frequency is, however, different in the two sexes. The estimate of this difference has been obtained calculating the relative risk (x) to contract the disease in the males, as compared to the females. Considering all the families together, such risk is x =5.8. However, x =16.7 in the 39 families with both healthy parents; x = 4.0 in the 45 families with the father affected; x = 2.9 in the families with the mother affected;3) though the analysis of segregation ratios would not allow, by itself, definitely excluding the hypothesis of a monofactorial recessive inheritance, many reasons lead to think of a possible case of « simulated mendelism »;4) the application of the methods of analysis proposed and worked out, in their theoretical basis, by Penrose (1953) and by Edwards (1960) for the genetic study of common diseases, makes much more probable the hypothesis that the genetic basis of duodenal ulcer is multifactorial. This does not exclude, however, that some particular environmental factors may increase the phenotypical correlation among the brothers — at least in the material under consideration.Therefore, the hypothesis now allowing the best interpretation of the observations is that duodenal ulcer could be controlled by a number of additive genes with a weak effect.