Over 5% of the cancer patient population may be radiation sensitive due to genetics, and the sensitive patients may be greatly overrepresented among patients with cancer therapy complications. These individuals include not only rare ataxia telangiectasia (AT) homozygotes with up to three-fold normal radiation sensitivity, but also far more numerous patients with slight radiosensitivity conjectured to be carriers of AT or to have another inherited mutagen sensitivity. Procedures may eventually be used to reliably determine patient tolerance for radiation and antineoplastic agents before initiation or completion of therapy, to have the therapy approach but not exceed the radiation tolerance of the individual patient's irradiated normal tissue. Such procedures could include study of patient's cultured normal cells (e.g., fibroblasts, marrow cells, or lymphocytes) in much the same way that patients' cultured tumor cells may eventually be widely used in the human tumor stem cell assay to predict which course of radiotherapy or chemotherapy should be most useful for treating a cancer. Studies with the normal cells could include cytotoxicity assays, serially determined accumulated genetic damage over the course of therapy, or Southern blot analysis to identify carriers of DNA repair mutations. Such studies could permit more aggressive radiotherapy of most patients due to the noninclusion of a sensitive subpopulation of patients, with less radiotherapy of the relatively few radiation sensitive patients. The patient's tumor cells should have inherited any radiation (or chemotherapy) sensitivity mutations present in the patient's normal cells, so reducing the radiotherapy dose to compensate for the more radiosensitive patients' sensitivity will not necessarily result in undertreatment of the tumor.