Mitochondria have bioenergetic properties that strongly suggest their involvement in the cellular formation of reactive oxygen species (ROS). Apparent confirmation of this process has come from work with isolated mitochondria, which have been shown to produce H(2)O(2) from dismutating superoxide radicals. Two different sites were reported to shuttle single electrons to oxygen out of the normal respiratory sequence. However, the mechanisms for ROS formation at these two sites are controversial. Arguments against mitochondrial ROS formation in the living cell are based on the fact that bioenergetic alterations may result from the mechanical removal of mitochondria from their natural environment. Furthermore, the invasive detection methods that are generally used may be inappropriate because of the possible interaction of the detection system with mitochondrial constituents. The use of non-invasive detection methods has proved that ROS formation does not occur unless changes in the physical state of the membrane are established. The aim of this commentary is to discuss critically the arguments in favor of mitochondria as the main intracellular source of ROS. The pros and cons of working with isolated mitochondria, as well as the detection methodology are carefully analyzed to judge whether or not the above assumption is correct. The conclusion that mitochondria are the main ROS generators in the cell contradicts the fact that ROS release was not observed. However, if electron flow from ubiquinol to the bc(1) complex is hindered due to changes in lipid fluidity, single electrons may transfer to dioxygen and produce H(2)O(2) via superoxide radicals.