1. A method of estimating absolute ionic permeability coefficients which does not depend on the use of impermeant substitutes is reported. 2. The method is based on a pump leak model of the Xenopus laevis oocyte membrane. The procedure consists of measuring, in the same experiment, the pump current and the currents generated under voltage clamp by the partial substitution of one or two ions at a time. For each experimental condition, the measured currents are substituted in a Goldman-Hodgkin-Katz type equation with two unknowns (the permeability coefficients). The set of equations thus generated enables the computation of all the ionic permeability coefficients. 3. The Xenopus oocyte membrane (stages IV and V, Dumont, 1972) has been found to be permeable to conventional ion substitutes such as N-methyl-D-glucamine (NMG), sulphate, isethionate and gluconate. 4. The values for sodium, potassium and chloride permeability coefficients obtained from sixty-eight pooled experiments were, respectively, 5.44, 17.41 and 1.49 x 10(-8) cm s-1. 5. The diffusional currents for sodium, potassium and chloride computed from the experiments referred to above were, respectively, -1.16, 0.69 and -0.038 microA cm-2. 6. A stoichiometry of the Na+-K+ pump exchange of 3/1.8 was computed. 7. The intracellular concentrations of sodium, potassium and chloride ions, as determined by ion-selective microelectrodes, were, respectively, 10.1 +/- 0.66 mM (n = 12), 109.5 +/- 3.3 mM (n = 13) and 37.7 +/- 1.18 mM (n = 19), corresponding to equilibrium potentials of 61, -95 and -28 mV. 8. Since chloride is not at equilibrium across the membrane, we propose that there is an inward uphill Cl- transport.