Mitochondria are often introduced as the ATP-producing “powerhouses” of eukaryotic (nucleus-bearing) cells, but they fulfill essential roles in a number of other cell processes, including biosyntheses, programmed cell death, and the assembly of iron-sulfur clusters, to name just a few. Mitochondria are always surrounded by two membranes, and most mitochondria, but not all, contain their own DNA, which is an evolutionarily reduced bacterial chromosome. Since the early 1900s, mitochondria were suspected to have arisen through an endosymbiosis — one cell coming to live within another. By the 1970s, the existence of DNA in mitochondria and the overall similarity between mitochondrial ATP-producing biochemistry and that in free-living bacteria provided strong evidence in favor of that view. There is no longer any doubt that mitochondria arose through endosymbiosis, but there is currently a plurality of ideas about the kind of bacterium the ancestral mitochondrial endosymbiont was, the nature of the host that acquired the endosymbiont, and the nature of the initial symbiotic interactions that associated the host and the endosymbiont in their fateful encounter.