Various recent investigations relevant to the study of aging mechanisms have recently found that increases in longevity during dietary restriction can occur together with lack of decreases or even increases in O2 consumption. This is frequently interpreted as contradictory with the mitochondrial free radical theory of aging. But this is based on the erroneous assumption that increasing O2 consumption must increase the rate of mitochondrial oxygen radical generation. Here it is shown that the opposite occurs in many important situations. Strong decreases in absolute and relative (per unit of O2 consumed) mitochondrial oxygen radical production occur during aerobic exercise bouts, chronic exercise training, and hyperthyroidism, and notably, during dietary restriction. Mitochondrial oxygen radical generation is also lower in long-lived birds than in short-lived mammals of similar body size and metabolic rate. Total rates of reactive oxygen species generation can also vary between tissues in a way not linked to their differences in oxygen consumption. All this indicates that mitochondrial reactive oxygen species (ROS) production is not a simple byproduct of mitochondrial respiration. Instead, it is regulated independently of O2 consumption in many different physiologic situations, tissues, and animal species. Thus, the apparently paradoxical increases in O2 consumption observed in some models of dietary restriction do not discredit the mitochondrial free radical theory of aging, and they can further strengthen it.