Genetico-statistical analysis was made to check various hypotheses of the tendency to multiple birth inheritance. The material involved was comprised of 115 MZ and 228 DZ twin families burdened by recurrent cases of multiple births in their genealogy. Test data included 516 single birth probands, 5 from which had twins among sibs; this results in p = 0,97% for the evaluation of population frequency of the "affected" couples having twins). Vienberg proband method was applied to check monogenous-autosomal model and Edward & Smith approximating formulae to check additive-polygenous model with liminal results (manifestation). It is shown: 1) that the degree of genetical determination of MZ and DZ twinning is approximately the same for both multiple birth types; H-61 and 53% respectively; 2) in MZ twinning both mother's and father's genotypes perform as multiple birth factors; H-76% for the group of mother's sisters and 64% for that of father's brothers; 3) in case of DZ twinning mother's genotype is much more valid as a multiple birth factor as compared to the father's one; H-68% for the group of mother's sisters and H-25% for father's brothers; 4) at least some genetical factors, involved in multiple birth determination, are common for MZ and DZ twins; the rate of DZ twinning (of different sexes) among sibs of parents of MZ twins is reliable and more than 5-fold increases that in common population. It is suggested that the contradiction of literary data on multiple birth genetics is due to unadequate methods in many early investigations: calculations have been carried out on the basis of twin birth rate, and not on the rate of "affected" (couples having twins); differential Veinberg's method has been used, which is adequate in populational analysis and is unsuitable for genealogical studies for the estimation of MZ and DZ twinning frequency; cases of "sporadic" multiple birth have not been excluded from summary family material. On the basis of the authors' and literary data it is suggested also that the number of main genetic factors determining the tendency to multiple births is more than 2 (probably 3) and does not exceed 5, and their interaction approximated by oligenic-complementary model, which does not exclude the presence of genocopying loci in a total system.