Summary. Microcapillary electrodes filled with a variety of salt solutions, including 1 M KC1, have been used to measure the membrane potentials and resistances of spherical cells from the mycelial fungus Neurospora (cell diameters 15-25 gm, cell volumes 3-8 pl). During impalements with electrodes containing 0.3-1.0 M KC1, membrane potential and resistance decayed over a period of 3-10 min. In contrast, electrodes filled with 0.1 M KC1 gave stable membrane potentials of -180 mV and membrane resistivities of 40 k~ cm 2, values comparable to earlier results from the fungal hyphae.Salt leakage from 1.0 M KCl-filled electrodes (tip diameters 0.2-0.3 pro, resistances 50-75M~2) occurred at rates of 4~5 fmol sec-1, as indicated by direct intracellular measurements with ion-sensitive microelectrodes. Depending on cell size, such leakage rates could elevate cytoplasmic KC1 content at initial rates of 30 170 mM min -1, and actual values as high as 70 mM min-1 were observed. Salt leakage and changes in cytoplasmic KC1 concentration were reduced five-to sevenfold when impalements were made with electrodes containing 0.1 M KC1.The effects on cell membrane parameters of salt leakage from microelectrodes could be attributed to chloride ions. Substitution of the KC1 electrolyte with half-molar K2SO4 or NazSO4 and molar concentrations of K-and Na-MES [potassium and sodium 2-(N-morpholino)ethanesulfonate] gave stable membrane potentials in excess of -200 mV and membrane resistivities greater than 50 k~ cm 2, while the permeant anions NO~-and SCN-depressed the membrane parameters in a manner similar to that observed with 1 M KC1. Furthermore, modest elevation of cytoplasmic chloride concentration (below ca. 50 mM) affected both membrane potential and resistance in direct proportion to the concentration, and could be quantitatively described by the Constant Field Theory with a fixed membrane permeability (Pc1 ~ 4 x 10-8 cm sec-1). Higher cytoplasmic chloride levels produced a collapse of the membrane resistance and drastic depolarization in a fashion requiring large changes of membrane permeability.At least for cells with volumes of 10 pl or less, the standard practice of filling electrodes with 1 or 3 M KC1 should be abandoned. Half-molar (and lower) concentrations of K2SO 4 or Na2SO 4 are suggested as satisfactory replacements.