Conditions have been established in which three phases of respiration-dependent H+ transfer occur in preparations of rat liver mitochondria and their relation to permeability, energy coupling, and ultrastructure is described. When inner membrane structure is extensively damaged (Phase I) oxidative phosphorylation is uncoupled and no apparent barrier to H+ transfer exists. Such mitochondria show a respiration-dependent uptake or consumption of H+. H+ uptake was correlated by dual-beam spectrophotometric measurements with cytochrome oxidation, hence this H+ transfer arises from electron transfer. Mitochondria whose inner membranes are structurally impaired by either aging, or treatment with small amounts of Triton, or uncoupling agents (Phase II) show a diminished ADP O ratio and respiratory control index as compared to untreated controls. They also evince increased permeability to protons in H+ pulse experiments. In O2 pulse experiments a H+ O ratio about 4 occurs when succinate or choline chloride are substrates. This residual production of H+ was largely independent of exogenous monovalent (K+, Na+) or divalent (Ca++ and Mg++) cations but a cation/H+ exchange arising from endogenous monovalent ions occurs that is sufficient to account for the observed proton transfer. Hence H+ transfer in structurally damaged mitochondria arises from cation uptake and not necessarily from vectorial H+ transfer predicted by the chemiosmotic hypothesis. Fresh, tightly coupled mitochondria (Phase III) show a respiration-dependent H+ uptake whose rate and magnitude depends on endogenous cofactors for oxidative phosphorylation. These mitochondria are characterized by a limited permeability to H+ ions and a relatively high ADP O ratio and respiratory control index. Thus undamaged rat liver mitochondria do not develop the respiration-dependent H+ gradients predicted by the chemiosmotic hypothesis. It is concluded that changes in mitochondrial structure control H+ transfer. © 1969.