N a/K/2Cl cotransport carrier plays an important role in fluid absorption and secretion in many epithelial tissues. The role of the carrier, however, in mammalian choroidal cerebrospinal fluid (CSF) production has been controversial. We used ventriculo-cisternal perfusion (VCP) labeled with blue dextran with or without bumetanide and measured choroidal CSF production in anesthetized, and paralyzed, mechanically ventilated dogs. During 3 h of VCP, mean intracerebroventricular and arterial pressures, PaCO2, pH, [HCO3-], and serum osmolality remained normal in both groups (n = 9 in each group). Beginning 90 min after the start of VCP, choroidal CSF production was measured every 15 min. In group I (control group), values for CSF production (means +/- SD) were 49 +/- 20, 49 +/- 21, 51 +/- 21, 51 +/- 23, 48 +/- 20, 56 +/- 24, and 48 +/- 20 microliters/min, at 90, 105, 120, 135, 150, 165, and 180 min, respectively. These values did not differ significantly from each other. In group II (bumetanide group), after baseline control CSF production had been determined at 90 and 105 min, bumetanide (10(-4) mol/liter) was added to VCP. Mean values for CSF production were 54 +/- 15 and 52 +/- 17 microliters/min before, and 39 +/- 25, 34 +/- 19, 28 +/- 10, 30 +/- 17, and 30 +/- 18 microliters/min after addition of bumetanide at 90, 105, 120, 135, 150, 165, and 180 min, respectively. Comparing the two groups, baseline values for CSF production measured at 90 and 105 min did not differ significantly. After addition of bumetanide (group II), however, decrements in CSF production varied from 30 +/- 27% at 120 min to 47 +/- 14% at 150 min, which were significantly different from changes in group I. The results of this study indicate that NaCl cotransport carrier is involved in secretion of CSF in dogs, and inhibition of the transporter results in approximately 50% reduction in CSF production.