The basic premise of the thrifty gene hypothesis is that certain populations may have genes that determine increased fat storage, which in times of famine represent a survival advantage, but in a modern environment result in obesity and type 2 diabetes. The concept finds support in a unique animal model (Psammomys obesus) as well as among high type 2 diabetes susceptibility populations, such as North American Indians and South Pacific islanders. However, in some developing communities (e.g., Black South Africans) the thrifty phenotype hypothesis of perinatal malnutrition causing beta-cell dysfunction seems a better explanation, but this remains a contentious issue. Several genes have already been identified as candidates for the thrifty genotype, including those encoding proteins of the insulin-signaling and leptin pathways, as well as intermediary fat metabolism. Particular interest lies in the peroxisome-proliferator activated receptors. An innovative approach might be to focus on the "mirror image" of the thrifty genotype-congenital lipoatrophic diabetes mellitus, whose molecular defect remains enigmatic. We conclude that the genetic basis of the thrifty genotype probably derives from the multiplicative effects of polymorphisms at several sites mentioned above, rather than a single regulatory abnormality.