1. A small electrical potential difference (541 +/- 48 muV, aqueous side negative) across rabbit corneal endothelium has been recently found. Its dependence on ambient [Na(+)], [K(+)], [H(+)] and metabolic and specific inhibitors was examined.2. Changes in concentration of the ions above either were known or were presently shown to affect the rate of fluid transport across this preparation (normal value: 5.2 +/- 0.4 mul./hr.cm(2)). Ionic concentration changes were also found here to influence potential difference in the same way as fluid transport. In the cases tested, the effects on both fluid transport and potential difference were reversible.3. Fluid transport and potential difference were both decreased or abolished in absence of Na(+), K(+) and HCO(3) (-), and when [H(+)] was decreased. Fluid transport and potential difference were saturable functions of [HCO(3) (-)] and half-saturation occurred in both cases at about 13 mM-HCO(3) (-). The potential difference was also a saturable function of [Na(+)] (half-saturation around 15 mM). There was a pH optimum for potential difference in the range 7.4-7.6. Lower pH values decreases the potential difference and the fluid transport, and a small (-100 muV) reversed potential was observed in the range of 5.3-5.5.4. Total replacement of Cl(-) by HCO(3) (-) or SO(4) (2-) produced no impairment on either fluid transport or potential difference.5. Carbonic anhydrase inhibitors (ethoxyzolamide 10(-5) or 10(-4)M and benzolamide 10(-3)M) produced a 40-60% decrease in the rate of fluid pumping. In contrast, ethoxyzolamide 10(-4)M or acetazolamide 10(-3)M did not produce any change in the potential difference. NaCN and Na iodoacetate (both 2 mM) eliminated the potential difference in 1-1.5 hr while in controls it lasted for 5-6 hr.6. Ouabain (10(-5)M) abolished the potential difference in less than 10 sec when added to the aqueous side, which suggests the existence of an electrogenic pump. This extremely fast time transient can be accounted for by the accessibility and simple geometry of the present monocellular layer. Ouabain abolished also the reversed potential difference observed at low pH.7. The data are interpreted in terms of a scheme similar to that advanced for other epithelia and in which (a) H(+) would be pumped into the intercellular spaces, while Na(+) and CO(2) would enter into the cells, and (b) Na(+) would be subsequently pumped into the aqueous humour, producing as a result the fluid movement observed. The actual origin of the potential difference is further discussed in terms of two contrasting possibilities: (i) one or more electrogenic pumps, and (ii) a neutral pump which would create a diffusion potential across ;leaky' intercellular junctions.