The early-life environment, in particular maternal diet during pregnancy, influences a wide range of organs and systems in adult offspring. Mounting evidence suggests that developmental programming can also influence health and disease in grand-offspring. Transgenerational effects can be defined as those persisting into an F2 generation, where the F0 mother experiences suboptimal diet during her pregnancy. In this review, we critically examine evidence for transgenerational developmental programming effects in human populations, focusing on metabolic and reproductive outcomes. We discuss evidence from historical cohorts suggesting that grandchildren of women exposed to famine and other dietary alterations during pregnancy may experience increased rates of later health complications than their control counterparts. The methodological difficulties with transgenerational studies in human cohorts are explored. In particular, the problems with assessing reproductive outcomes in human populations are discussed. In light of the relative paucity of evidence available from human cohorts, we consider key insights from transgenerational experimental animal models of developmental programming by maternal diet; data are drawn from a range of rodent models, as well as the guinea-pig and the sheep. The evidence for different potential mechanisms of transgenerational inheritance or re-propagation of developmental programming effects is evaluated. Transgenerational effects could be transmitted through methylation of the gametes via the paternal and maternal lineage, as well as other possible mechanisms via the maternal lineage. Finally, future directions for exploring these underlying mechanisms further are proposed, including utilizing large, well-characterized, prospective pregnancy cohorts that include biobanks, which have been established in various populations during the last few decades.