AbstractA number of human genetic diseases have come to be described as being defective in DNA repair. The minimum criterion on which this assignment is based is hypersensitivity to the clastogenic or lethal action of specific DNA damaging agents. In one disease, xeroderma pigmentosum, the molecular evidence for a defect in DNA repair is unequivocal. This condition then acts as a model for dissecting others. For the other diseases the formal evidence for defects in repair is less secure or even lacking. The evidence for repair in each disease is assembled together with any methods that have been used to support the differential diagnosis or for prenatal diagnosis. Attempts to clone human DNA repair genes are in hand and may provide the necessary evidence to decide if all the putative DNA repair defective diseases are genuine.Neoplastic disease and neurological degeneration together with immune defects are frequent clinical features linking this set of diseases, suggesting that effective DNA repair may be important in many aspects of human health.