Phlorizin-sensitive currents mediated by a Na-glucose cotransporter were measured using intact or internally perfused Xenopus laevis oocytes expressing human SGLT1 cDNA, Using a two-microelectrode voltage clamp technique, measured reversal potentials (V-r) at high external alpha-methylglucose (alpha MG) concentrations were linearly related to In[alpha MG](o), and the observed slope of 26.1 +/- 0.8 mV/decade indicated a coupling ratio of 2.25 +/- 0.07 Na ions per alpha MG molecule, As [alpha MG](o) decreased below 0.1 mM, V-r was no longer a linear function of In[alpha MG](o), in accordance with the suggested capacity of SGLT1 to carry Na in the absence of sugar (the ''Na leak''). A generalized kinetic model for SGLT1 transport introduces a new parameter, K-c, which corresponds to the [alpha MG](o) at which the Na leak is equal in magnitude to the coupled Na-alpha MG flux. Using this kinetic model, the curve of V-r as a function of In[alpha MG](o) could be fitted over the entire range of [alpha MG](o) if K-c is adjusted to 40 +/- 12 mu M. Experiments using internally perfused oocytes revealed a number of previously unknown facets of SGLT1 transport. In the bilateral absence of alpha MG, the phlorizin-sensitive Na leak demonstrated a strong inward rectification. The affinity of alpha MG for its internal site was low; the K-m was estimated to be between 25 and 50 mM, an order of magnitude higher than that found for the extracellular site. Furthermore, V-r determinations at varying alpha MG concentrations indicate a transport stoichiometry of 2 Na ions per alpha MG molecule: the slope of V-r versus In[alpha MG](o) averaged 30.0 +/- 0.7 mV/decade (corresponding to a stoichiometry of 1.96 +/- 0.04 Na ions per alpha MG molecule) whenever [alpha MG](o) was higher than 0.1 mM. These direct observations firmly establish that Na ions can utilize the SGLT1 protein to cross the membrane either alone or in a coupled manner with a stoichiometry of 2 Na ions per sugar molecule.