The cytoplasmic membrane ionic current of cells of Rhodobacter capsulatus, washed to lower the endogenous K+ concentration, had a non‐linear dependence on the membrane potential measured during photosynthetic illumination. Treatment of the cells with venturicidin, an inhibitor of the H+‐ATP synthase, increased the membrane potential and decreased the membrane ionic current at values of membrane potential below a threshold. The addition of K+ or Rb+, but not of Na+, led to an increase in the membrane ionic current and a decrease in the membrane potential in either the presence or absence of venturicidin. Approximately 0.4 mM K+ or 2.0 mM Rb+ led to a half‐maximal response. At saturating concentrations of K+ and Rb+, the membrane ionic currents were similar. The membrane ionic currents due to K+ and Rb+ were not additive. The K+‐dependent and Rb+‐dependent ionic currents had a non‐linear relationship with membrane potential: the alkali cations only increased the ionic current when the membrane potential lay above a threshold value. The presence of 1 mM Cs+ did not lead to an increase in the membrane ionic current but it had the effect of inhibiting the membrane ionic current due to either K+ or Rb+. Photosynthetic illumination in the presence of either K+ or Rb+, and weak acids such as acetate, led to a decrease in light‐scattering by the cells. This was attributed to the uptake of potassium or rubidium acetate and a corresponding increase in osmotic strength in the cytoplasm. The addition of NH4+ also led to an increase in membrane ionic current and to a decrease in membrane potential (half‐maximal at 2.0 mM NH4+). The relationship between the NH4+‐dependent ionic currents and the membrane potential was similar to that for K+. The NH4+‐dependent and K+‐dependent ionic current were not additive. However, illumination in the presence of NH4+ and acetate did not lead to significant light‐scattering changes. The NH4+‐dependent membrane ionic current was inhibited by 1 mM Cs+ but not by 50 μM methylamine. It is proposed that the K+‐dependent membrane ionic current is catalysed by a low‐affinity K+‐transport system such as that described in Rb. capsulatus [Jasper, P. (1978) J. Bacteriol. 133, 1314–1322]. The possibility is considered that, as well as Rb+, this transport system can also operate with NH4+. However, in our experimental conditions NH4+ uptake is followed by NH3 efflux. This process can compete effectively with uptake of weak acids for the trans‐membrane pH gradient. The result of electrophoretic uptake of NH4+ and the efflux of NH3 from the cell is a ‘futile’ partial dissipation of both the trans‐membrane pH gradient and membrane potential components of the protonmotive force. The increase in K+‐dependent (and Rb+‐and NH4+‐dependent) membrane ionic current is not instantaneous with the increase in membrane potential at the onset of illumination. This is taken to indicate a slow conformational change (t1/2= 120 ms) in the activation of the transporter by membrane potential. A membrane potential exists even in darkened anaerobic cells of Rb. capsulatus. As recently shown by measurements of the uptake of tetraphenylphosphonium cation in Rb. sphaeroides [Abee, T., Hellingwerf, K. J., Konings, W. N. (1988) J. Bacteriol. 170, 5647–5653] this potential was decreased in the presence of K+. In Rb. capsulatus the depression of this potential by K+ was linearly related to the logarithm of the K+ concentration in the external medium. Copyright © 1990, Wiley Blackwell. All rights reserved