We have demonstrated previously active Cl and K secretion by the cortical collecting tubule (CCT). The aim of these studies was to determine whether a component of Cl and K secretion is coupled. Such coupling has been observed in the rat distal nephron but K secretion in the rabbit CCT is believed to be strictly conductive. We measured the rate of K secretion in three protocols. In the first, bath bumetanide did not affect K secretion but K secretion (-JK; in pmol.mm-1.min-1) decreased from 12.7 +/- 2.2 to 8.28 +/- 1.2 when gluconate replaced bath Cl (P less than 0.01), whereas transepithelial voltage (VT) and lumen-to-bath 22Na flux (JNal----b) were unchanged. In the second, bath Cl removal stimulated significant K absorption (JK, from -0.19 +/- 0.97 to 2.03 +/- 0.57) only in the presence of a lumen-to-bath Cl gradient (Na-free perfusate). This identical maneuver had no effect on JK in the absence of luminal Cl (NaCl-free perfusate). In the third, reducing luminal [Cl] from 112 to 5 mM stimulated K secretion from 13.6 +/- 3.1 to 20.1 +/- 3.2 (P less than 0.01) without affecting VT or net Na transport. We conclude the following: the CCT possesses a Cl-dependent K secretory mechanism that can be influenced by the transepithelial Cl gradient independent of VT and JNal----b. Reducing luminal [Cl] stimulates K secretion and K secretion can be reversed to K absorption by reversal of the ambient Cl gradient. These data imply a coupling between a component of K and Cl fluxes, consistent with the presence of KCl cotransport in the rabbit CCT.