We suggest that altered fetal growth and type 2 diabetes may be two phenotypes of the same genotype - in other words the 'thrifty phenotype' is the result of a 'thrifty genotype'. Supporting this there is strong evidence that paternal factors and, therefore, genes influence fetal growth and that these paternal genes affecting fetal growth may also alter diabetes risk. Further study is needed to determine whether common gene variants can explain the association between reduced birth weight and increased risk of type 2 diabetes. If the genetic hypothesis is true, common diabetes genes are likely to have subtle effects on insulin secretion and/or action and, therefore, subtle effects on fetal growth. Large cohorts of infants and their parents will be required - probably in the region of thousands rather than hundreds - to identify gene variants that may explain the association between reduced birth weight and increased risk of type 2 diabetes. All previously described associations between birth weight and type 2 diabetes have required many hundreds of subjects and it is likely that the geneticists and the 'programmists' are trying to identify very subtle physiological effects.