Submicromolar to picomolar lower detection limits have recently been obtained with various polymer membrane ion-selective electrodes by minimizing biases due to ion fluxes through the membrane. For the best performance, the compositions of the membrane and inner solution should be optimized for each application. Given the number of parameters to be adjusted, it has been difficult to find the best parameters for a target sample. In this paper, a much simplified and more practical steady-state model of zero-current ion fluxes is derived, which is based on measurable parameters. The model allows one to predict achievable lower detection limits for a membrane with given selectivities. It can also be used to predict the optimal composition of the inner filling solution for the measurement of samples with a known, typical ionic background. Selectivity coefficients of monovalent and divalent analyte ions required for desired detection limits in drinking water are calculated. As an application of the proposed general recipe, a silver-selective electrode is developed on the basis of the ionophore O,O''-bis[2-(methylthio)ethyl]-tert-butylcalix[4]arene. With the predicted optimal composition of the inner electrolyte, its lower detection limit is found to be 10(-9) M or 100 ppt Ag+ with an ionic background of 10(-5) M LiNO3, which is very close to the expected value.