Here we explore whether there is any evidence that the rapid development of the obesity epidemic in emerging nations, and its unusual coexistence with malnutrition, may have evolutionary origins that make such populations especially vulnerable to the obesogenic conditions accompanying the nutrition transition. It is concluded that any selection of so-called 'thrifty genes' is likely to have affected most races due to the frequency and ubiquity of famines and seasonal food shortages in ancient populations. Although it remains a useful stimulus for research, the thrifty gene hypothesis remains a theoretical construct that so far lacks any concrete examples. There is currently little evidence that the ancestral genomes of native Asian or African populations carry particular risk alleles for obesity. Interestingly, however, there is evidence that a variant allele of the FTO gene that favors leanness may be less active in Asians or Africans. There is also some evidence that Caucasians may be less prone to developing type 2 diabetes mellitus than other races suggesting that there has been recent selection of protective alleles. In the near future, recently developed statistical methods for comparing genome-wide data across populations are likely to reveal or refute the presence of any thrifty genes and might indicate mechanisms of vulnerability.