Recent works on the structure and the function of cytochrome-c oxidase are reviewed. The subunit composition of the mitochondrial enzyme depends on the species and is comprised of between 5 and 13 subunits. It is reduced to 1 to 3 subunits in prokaryotes. The complete amino acid composition has been derived from protein sequencing. Gene sequences are partially known in several eukaryote species. Metal centers are only located in subunits I and II. The mitochondrial cytochrome-c oxidase is Y-shaped; the arms of the Y cross the inner membrane, the stalk protrudes into the intermembrane space. The bacterial enzyme has a simpler, elongated shape. A number of data have been accumulated on the subunit topology and on their location within the protein. All available spectrometric techniques have been used to investigate the environment of the metal centers as well as their interactions. From the literature, attention must be paid to what may be considered or not as an active form. The steady improvement of the instrumentation has yielded evidence for different kinds of heterogeneities which could reflect the in vivo situation. The 'pulsed' and 'resting' conformers have been well characterized. The 'oxygenated' form has been identified as a peroxide derivative of the fully oxidized cytochrome-c oxidase. The mammalian enzyme has been isolated in fully active monomeric form which does not preclude the initially suggested dimeric behavior in situ. The role of the lipids is still largely investigated, mainly through reconstitution experiments. Kinetic studies of electron transfer between cytochrome c and cytochrome-c oxidase lead to a single catalytic site model to account for the multiphasic kinetics. Results related to the low temperature investigation of the intermediate steps in the reaction between oxygen and cytochrome-c oxidase received a sound confirmation by the resolution of compound A at room temperature. It is also pointed out that the so-called mixed valence state might not be a transient state in the catalytic reduction of oxygen. The functioning of cytochrome-c oxidase as a proton pump has been supported by a number of experimental results. Subunit III would be involved in this process. The redox link to the proton pump has been suggested to be at the Fea-CuA site. The molecular mechanism responsible for the proton pumping is still unknown.