The mechanism of cation translocation by the Na,K-ATPase was investigated by cysteine scanning mutagenesis and measurements of accessibility through exposure to cysteine reagents. In the native protein, accessible residues were found only at the most extracellular residues of the 5th and 6th transmembrane segments (TMS) and the short loop between them. However, after modification by palytoxin a number of residues became accessible along the whole length of the 5th TMS and in the outer half of the 6th TMS, showing the contribution of each of these segments to the "channel" formed by the palytoxin-transformed Na,K-pump. Assuming that this structure is similar in the native and the palytoxin-transformed pump, our data allow us to determine the residues lining the cation pathway from the extracellular solution to their binding sites. A critical position in the 5th TMS contains a lysine conserved in all known nonelectrogenic H,K-ATPases, and a serine in all known electrogenic Na,K-ATPase sequences. Wild-type or mutant Na,K-or H,K-ATPase a subunits were coinjected with the Bufo beta2 subunit in Xenopus oocytes and Rb(86) uptake and electrophysiological measurements were performed. An electrogenic activity was recorded for the H,K-ATPase mutants in which the positively charged lysine had been replaced by neutral or negatively charged residues, while nonelectrogenic transport was observed with the S(782)R mutant of the Na,K-ATPase. The presence or the absence of a positively charged residue at the S(782) position appears to be critical for the stoichiometry of cation exchange.