On illumination, chromatophores of Rhodospirillum rubrum took up hydrogen ions from the medium. Hydrogen ion uptake was inhibited or reversed by uncoupling agents and by Triton‐X114. Addition of valinomycin in the dark to chromatophores suspended in a medium of low KCl concentration gave rise to a loss of potassium ion from the chromatophores, and a slow H+‐uptake. Illumination gave rise to a further efflux of potassium ion, and a stimulated H+‐uptake. The light‐induced potassium ion efflux increased in extent with increasing external KCl concentration. At high KCl concentrations hydrogen ion efflux occurred on addition of valinomycin in the dark, and a greatly stimulated initial rate of H+‐uptake on illumination was observed. Addition of gramicidin to chromatophores in the dark gave rise to effects similar to those observed with valinomycin. On illumination, the uptake of hydrogen ion was somewhat inhibited, and rapidly reversed in the dark. Addition of nigericin in the dark to chromatophores suspended in a medium of low KCl concentration gave rise to a rapid efflux of potassium ion, and an uptake of hydrogen ion which was nearly stoichiometric in rate and extent. On illumination an inhibited H+‐uptake occurred, together with an uptake of potassium ion from the medium. The inhibition of H+‐uptake, and the extent of K+‐uptake were dependent on nigericin concentration and KCl concentration. The sum of the extent of K+‐uptake and the extent of H+‐uptake was approximately the same in all cases, indicating that an exchange of hydrogen ion within the chromatophores for potassium ion in the medium had occurred. Nigericin gave rise to hydrogen ion efflux from chromatophores suspended in media of high KCl concentration when added in the dark or light. The efflux observed in the light was more extensive and represented a reversal of hydrogen ion uptake. The KCl concentration at which nigericin caused no pH change was shifted from about 4 mM in the dark to about 0.4 mM in the light. The pH gradient across the chromatophore membrane was calculated to be approximately 1 pH unit. Valinomycin inhibited nigericin induced K+‐uptake in the light. At high KCl concentrations nigericin inhibited H+‐uptake in the absence or presence of valinomycin. No marked inhibition of photophosphorylation by nigericin or valinomycin alone was observed even at high KCl concentrations. However, when both antibiotics were present together in the reaction medium, inhibition was observed which was dependent on the KCl concentration. In the absence of added KCl, no inhibition occured but at 10 mM KCl phosphorylation was 86%, and at 100 mM, 93% inhibited. Similar results were obtained when phosphorylation was coupled to electron flow in uninhibited chromatophores, or through N‐methyl phenazonium methosulphate in 2‐heptyl‐4‐hydroxy quinoline N‐oxide inhibited chromatophores.Similar effects were observed when dianemycin was used instead of nigericin. The mechanism of action of the antibiotics tested is discussed. It is suggested that their mode of action on chromatophore membranes is the same as that shown for artificial membrane systems. In particular, it is proposed that nigericin and dianemycin are able to catalyse a cation/cation or cation/hydrogen ion exchange across lipid barriers, which is closely coupled and electrically neutral.