The effects of ion substitution, inhibitors and variations in transmural p.d. on the movements of sodium and chloride across an in vitro preparation of rat gastric mucosa have been studied. The tissue maintained net steady‐state transport of sodium in the mucosal‐to‐serosal direction in the absence of transmural gradients of electrochemical potential. Sodium transport was independent of the presence of chloride, and was abolished by 1 X 10(‐5) M‐amiloride. The inhibitor produced a decrease in short‐circuit current equivalent to the depression of sodium transport, indicating that the sodium transport process was electrogenic. Variations in transmural p.d. showed that the sodium transport process included two components: one that varied with p.d. and one that was independent of it. These findings have been interpreted in terms of a system for sodium transport composed of three components: two rate‐limiting entry mechanisms at the apical membrane, one of which can be represented as a conductive channel for sodium diffusion and the other as a neutral process possibly a sodium‐hydrogen exchanger, and a voltage‐independent pump at the basolateral membrane analogous to the constant‐current pump models described in some other epithelia. The tissue maintained a net secretory movement of chloride in the short‐circuited condition. The process responsible for net transport of chloride could be resolved into two components: one that was sodium dependent, electrogenic, and abolished by 8 X 10(‐3) M‐acetazolamide, and one that was independent of the presence of sodium, electrically silent and abolished by 5 X 10(‐4) M‐SITS (4‐acetamido‐4'‐isothiocyano‐2,2'‐disulphonic acid stilbene). Both components of the chloride transport process varied with p.d. These findings were interpreted in terms of a system of three components: two entry mechanisms at the basolateral membrane including a coupled sodium‐chloride influx process and a chloride‐bicarbonate exchanger in parallel, and a rate‐limiting conductive channel at the apical membrane. In addition, the studies on the effects of variations in transmural p.d. on chloride fluxes revealed a symmetrical voltage‐independent component, dependent on the presence of chloride in the trans compartment, and it was suggested that this component may reflect the presence of a chloride‐chloride exchange mechanism.