An early feature of Alzheimer's disease (AD) is region-specific declines in brain glucose metabolism. Unlike other tissues in the body, the brain does not efficiently metabolize fats; hence the adult human brain relies almost exclusively on glucose as an energy substrate. Therefore, inhibition of glucose metabolism can have profound effects on brain function. The hypometabolism seen in AD has recently attracted attention as a possible target for intervention in the disease process. One promising approach is to supplement the normal glucose supply of the brain with ketone bodies (KB), which include acetoacetate, beta-hydroxybutyrate, and acetone. KB are normally produced from fat stores when glucose supplies are limited, such as during prolonged fasting. KB have been induced both by direct infusion and by the administration of a high-fat, low-carbohydrate, low-protein, ketogenic diets. Both approaches have demonstrated efficacy in animal models of neurodegenerative disorders and in human clinical trials, including AD trials. Much of the benefit of KB can be attributed to their ability to increase mitochondrial efficiency and supplement the brain's normal reliance on glucose. Research into the therapeutic potential of KB and ketosis represents a promising new area of AD research.