The main message of this review can be summarized as follows: aging and longevity, as complex traits having a significant genetic component, likely depend on a number of nuclear gene variants interacting with mtDNA variability both inherited and somatic. We reviewed the data available in the literature with particular attention to human longevity, and argued that what we hypothesize for aging and longevity could have a more general relevance and be extended to other age-related complex traits such as Alzheimer's and Parkinson's diseases. The genetics which emerges for complex traits, including aging and longevity, is thus even more complicated than previously thought, as epistatic interactions between nuclear gene polymorphisms and mtDNA variability (both somatic and inherited) as well as between mtDNA somatic mutations (tissue specific) and mtDNA inherited variants (haplogroups and sub-haplogroups) must be considered as additional players capable of explaining a part of the aging and longevity phenotype. To test this hypothesis is one of the main challenge in the genetics of aging and longevity in the next future.