Bidirectional transepithelial K+ flux measurements across 'high-resistance' epithelial monolayers of MDCK cells grown upon millipore filters show no significant net K+ flux. Measurements of influx and efflux across the basal-lateral and apical cell membranes demonstrate that the apical membranes are effectively impermeable to K+. K+ influx across the basal-lateral cell membranes consists of an ouabain-sensitive component, an ouabain-insensitive component, an ouabain-insensitive but furosemide-sensitive component, and an ouabain- and furosemide-insensitive component. The action of furosemide upon K+ influx is independent of (Na+ - K+)-pump inhibition. The furosemide-sensitive component is markedly dependent upon the medium K+, Na+ and Cl- content. Acetate and nitrate are ineffective substitutes for Cl-, whereas Br- is partially effective. Partial Cl- replacement by NO3 gives a roughly linear increase in the furosemide-sensitive component. Na+ replacement by choline abolishes the furosemide-sensitive component, whereas Li+ is a partially effective replacement. Partial Na+ replacement by choline abolishes the furosemide-sensitive component, whereas Li+ is a partially effective replacement. Partial Na+ replacement with choline gives an apparent affinity of approximately 7 mM Na, whereas variation of the external K+ content gives an affinity of the furosemide-sensitive component of 1.0 mM. Furosemide inhibition is of high affinity (K1/2 = 3 micrometer). Piretanide, ethacrynic acid, and phloretin inhibit the same component of passive K+ influx as furosemide; amiloride, 4,-aminopyridine, and 2,4,6-triaminopyrimidine partially so. SITS was ineffective. Externally applied furosemide and Cl- replacement by NO3- inhibit K+ efflux across the basal-lateral membranes indicating that the furosemide-sensitive component consists primarily of K:K exchange.