1. The guinea-pig placenta perfused in situ via the umbilical circulation has been used to measure unidirectional fluxes of Na from mother to fetus, and in the reverse direction, with 24Na and 22Na. There was no significant difference between the two fluxes, each being 22 mumole.min-1. 2. Ouabain 10(-5) M in the perfusion fluid had no detectable effect on radioisotopic movements of Na in either direction. 3. Unidirectional fluxes of 42K in both directions were approximately equal at 1.7 mumole.min-1 mother fetus and 1.8 mumole.min-1 in the reverse direction, despite a K concentration of 3.4 mM on the maternal side and 5.0 mM on the fetal side of the placenta. 4. Extraction of 42K and 86Rb from the perfusion fluid was inhibited by 43% by 10(-5) M-ouabain in the fluid. This effect was largely due to a reduction of isotope uptake by the placental tissue. 5. The relative permeabilities of the placenta, mother to fetus, were Rb approximately K (3.2) greater than Na (1.0) greater than Li (0.55). 6. Under the experimental conditions, the electrical potential difference between perfusion fluid and maternal blood was 6 mV (fetus negative). It was shifted towards the positive by a low Na fluid. 7. The results suggest the presence of a dominant Na-K pump (active component towards mother) sited at the maternal-facing membrane of the syncytiotrophoblast together with a subsidiary pump oriented in the opposite direction and probably sited together with a subsidiary pump oriented in the opposite direction and probably sited at the fetal-facing membrane of the syncytiotrophoblast. 8. A high proportion of Na movement particularly towards the fetus is probably passive, occurring through water-filled spaces, whilst K movement is more dependent on active transport.