One of the most robust observations in the biology of aging is that caloric restriction (CR) extends life in a variety of species. Although CR results in a severalfold decrease in fat mass (FM), the role of fat on life extension was considered to be minimal. Two main reasons accounted for this belief. First, although increased FM is associated with changes in substrate oxidation and in glucose homeostasis, in part through the effects of free fatty acids (FFA) and glycerol, several studies have suggested that longevity is determined independent of FM. Second, CR has systemic effects on a range of functions including neurological, endocrine, reproductive, immunological and antineoplastic, none of which have been historically linked to fat. In the last few years, an explosion of evidence has demonstrated that fat tissue is a very active endocrine gland which secretes a variety of peptides (such as leptin and plasminogen activating inhibitor-1), cytokines (such as tumor necrosis factor), and complement factors (such as D, C3, and B). This is in addition to the presence of substrates, such as glycerol and FFA, which are stored and released by fat cells and are known to have a major role in hepatic and peripheral glucose metabolism. We propose that many of the systemic effects of CR can now be explained by the chronic effects related to decreased plasma levels of peptides, cytokines, complement factors, and substrates. In fact, all of the benefits of CR on the neuroendocrine system and those related to the improvement in glucose homeostasis can be attributed to decrease in adipose cells and their products. Other evidence from epidemiological data in human obesity supports the role of fat mass and its body distribution as a risk factor for morbidity and mortality in humans due to impaired glucose metabolism (similar to rodents), for cancer (similar to rodents), and for the development of atherosclerotic vascular disease (in humans). If all or most of the life-extending benefits of CR can be attributed to decreased fat stores, the expression of specific candidate proteins may be explored and manipulated in the search for the most powerful adipose-dependent signals that modulate life expectancy.