It is established that calorie restriction (CR) increases the resistance of cells to various stressors such as oxidative damage, excitotoxins, mercury and acetaminophen. Alternate day feeding (ADF) may confer greater stress resistance than daily CR of 30% or 40%. A recent study in three strains of mouse showed that a fast of 48 or 60 h prevented toxic effects due to administration of doses 2-4 times the maximum human dose of etoposide, a chemotherapy agent which acts through increased oxidative stress. In addition, mice inoculated with neuroblastoma survived longer when pretreated with fasting, then given high dose etoposide, as well as not exhibiting toxicity. This increased survival was construed as evidence of differential stress resistance between normal and cancer cells, the cancer cells being only partially protected by the pretreatment fast. In clinical practice, increased differential stress resistance could lead to the use of much higher doses of chemotherapy agents, and in the absence of toxicity, make it possible to repeat the treatment to kill residual cancer cells. Humans are unlikely to comply with a total fast of longer than 24 or 48 h, which may be insufficient to activate the same gene expression process. Based on published data we estimate that an optimal time period for development of stress resistance is 2-3 weeks when alternate day feeding is employed. Our previously published experience suggests that 2-3 weeks of alternate day modified fast in which subjects eat ad libitum one day and <20% of one's estimated caloric requirement the next will confer a similar stress resistance. Compliance with this diet is high and greater maintenance of body weight is feasible. We hypothesize that a pretreatment of 2-3 weeks with the alternate day modified fast will improve outcomes in cancer chemotherapy, decreasing morbidity and raising cure rates.