The committee of the United Nations for the Evaluation of Radiation Effects UNSCEAR as well as the International Commission on Radiological Protection ICRP up to now have derived a very low risk for hereditary diseases in humans from experiments in mice. They claim that there are no human data to refer to, and missing effects in the acute exposed Japanese A-bomb survivors are erroneously generalized to situations of chronic exposure. We made a compilation of findings about early deaths, congenital malformations, Downs syndrome, cancer and other effects, which were observed in humans after the exposure of parents. A few pointers are available from occupationally exposed groups, and much information can be drawn from studies in populations exposed by Chernobyl fallout and from the descendants of liquidators. Nearly all types of hereditary defects were found, which are to be expected ac-cording to our general knowledge of mechanism. We show that the official risk estimates are much too low.