Isolated small intestine of toad (Bufo bufo) was mounted on glass tubes for perfusion studies with oxygenated amphibian Ringer's solution containing glucose and acetate. Under open-circuit conditions (Vt = -3.9 +/- 1.8 mV, N = 14) the preparation generated a net influx of 134Cs+. The time course of unidirectional 134Cs+-fluxes was mono-exponential with similar rate constants for influx and outflux when measured in the same preparation. The flux-ratio was time invariant from the beginning of appearance of the tracers to steady state was achieved. Thus, just a single pathway, the paracellular pathway, is available for transepithelial transport of Cs+. From the ratio of unidirectional Cs+-fluxes the paracellular force was calculated to be, 18.2 +/- 1.5 mV (N = 6), which is directed against the small transepithelial potential difference. The paracellular netflux of cesium ions, therefore, is caused by solvent drag. The flux of 134Cs+ entering and trapped by the cells was of a magnitude similar to that passing the paracellular route. Therefore, independent of the convective flux of 134Cs+, every second 134Cs+ ion flowing into the lateral space was pumped into the cells rather than proceeding, via the low resistance pathway, to the serosal bath. It is thus indicated that the paracellular convective flow of 134Cs+ is driven by lateral Na+/K+-pumps. Transepithelial unidirectional 42K+ fluxes did not reach steady state within an observation period of 70 min, indicating that components of the fluxes in both directions pass the large cellular pool of potassium ions. The ratio of unidirectional 24Na+ fluxes was time-variant and declined from an initial value of 3.66 +/- 0.34 to a significantly smaller steady-state value of 2.57 +/- 0.26 (P < 0.001, N = 5 paired observations), indicating that sodium ions pass the epithelium both via the paracellular and the cellular pathway. Quantitatively, the larger ratio of paracellular Na+ fluxes, as compared to that of paracellular Cs+ fluxes, is compatible with convective flow of the two alkali metal ions through the same population of water-filled pores. With a new set of equations, the fraction of the sodium flux passing the basement membrane barrier of the lateral space that is recirculated through the cellular compartment is estimated. This fraction was, on average, 0.72 +/- 0.03 (N = 5). It is concluded that isotonicity of the transportate can be maintained by producing a hypertonic fluid emerging from the lateral space combined with reuptake of salt via the cells.