The most studied comparison of aging and maximum lifespan potential (MLSP) among endotherms involves the 7-fold longevity difference between rats (MLSP 5y) and pigeons (MLSP 35y). A widely accepted theory explaining MLSP differences between species is the oxidative stress theory, which purports that reactive oxygen species (ROS) produced during mitochondrial respiration damage bio-molecules and eventually lead to the breakdown of regulatory systems and consequent death. Previous rat-pigeon studies compared only aspects of the oxidative stress theory and most concluded that the lower mitochondrial superoxide production of pigeons compared to rats was responsible for their much greater longevity. This conclusion is based mainly on data from one tissue (the heart) using one mitochondrial substrate (succinate). Studies on heart mitochondria using pyruvate as a mitochondrial substrate gave contradictory results. We believe the conclusion that birds produce less mitochondrial superoxide than mammals is unwarranted. We have revisited the rat-pigeon comparison in the most comprehensive manner to date. We have measured superoxide production (by heart, skeletal muscle and liver mitochondria), five different antioxidants in plasma, three tissues and mitochondria, membrane fatty acid composition (in seven tissues and three mitochondria), and biomarkers of oxidative damage. The only substantial and consistent difference that we have observed between rats and pigeons is their membrane fatty acid composition, with rats having membranes that are more susceptible to damage. This suggests that, although there was no difference in superoxide production, there is likely a much greater production of lipid-based ROS in the rat. We conclude that the differences in superoxide production reported previously were due to the arbitrary selection of heart muscle to source mitochondria and the provision of succinate. Had mitochondria been harvested from other tissues or other relevant mitochondrial metabolic substrates been used, then very different conclusions regarding differences in oxidative stress would have been reached. © 2011 Montgomery et al.