From 2.5 to 2.0 billion years ago, atmospheric oxygen concentration [O2] rose thousands of times, leading to the first mass extinction. Reactive Oxygen Species (ROS) produced by the non-catalyzed partial reduction of O2 were highly toxic eliminating many species. Survivors developed different strategies to cope with ROS toxicity. At the same time, using O2 as the final acceptor in respiratory chains increased ATP production manifold. Thus, both O2 and ROS were strong drivers of evolution, as species optimized aerobic metabolism while developing ROS-neutralizing mechanisms. The first line of defense is preventing ROS overproduction and two mechanisms were developed in parallel: 1) Physiological uncoupling systems (PUS), which increase the rate of electron fluxes in respiratory systems. 2) Avoidance of excess [O2]. However, it seems that as avoidance efficiency improved, PUSs became less efficient. PUS includes branched respiratory chains and proton sinks, which may be proton specific, the mitochondrial uncoupling proteins (UCPs) or unspecific, the mitochondrial permeability transition pore (PTP). High [O2] avoidance also involved different strategies: 1) Cell association, as in biofilms or in multi-cellularity allowed gas-permeable organisms (oxyconformers) from bacterial to arthropods to exclude O2. 2) Motility, to migrate from hypoxic niches. 3) Oxyregulator organisms: as early as in fish, and O2-impermeable epithelium excluded all gases and only exact amounts entered through specialized respiratory systems. Here we follow the parallel evolution of PUS and O2-avoidance, PUS became less critical and lost efficiency. In regard, to proton sinks, there is fewer evidence on their evolution, although UCPs have indeed drifted in function while in some species it is not clear whether PTPs exist.