After reviewing the history of mitochondrial diseases, I follow a genetic classification to discuss new developments and old conundrums. In the field of mitochondrial DNA (mtDNA) mutations, I argue that we are not yet scraping the bottom of the barrel because: (i) new mtDNA mutations are still being discovered, especially in protein-coding genes; (ii) the pathogenicity of homoplasmic mutations is being revisited; (iii) some genetic dogmas are chipped but not broken; (iv) mtDNA haplotypes are gaining interest in human pathology; (v) pathogenesis is still largely enigmatic. In the field of nuclear DNA (nDNA) mutations, there has been good progress in our understanding of disorders due to faulty intergenomic communication. Of the genes responsible for multiple deletions and depletion of mtDNA, mutations in POLG have been associated with a great variety of clinical phenotypes in humans and to precocious aging in mice. Novel pathogenetic mechanisms include alterations in the lipid milieu of the inner mitochondrial membrane and mutations in genes controlling mitochondrial motility, fission, and fusion.