Under in vitro conditions the rat cecum transported HCO3- from the serosal to an unbuffered solution in contact with the mucosal side [Js----m = 7.12 +/- 0.18 mumol.cm-2.h-1 (n = 149)]. With reversed tissues, a significantly lower flux was obtained [Jm----s = 2.47 +/- 0.11 mumol.cm-2.h-1 (n = 42)]. Both fluxes were stable for several hours. Increasing the H+ gradient across the tissue for 60 min did not change either flux. Anoxia for 45 min reversibly reduced Js----m by 65 +/- 3% (n = 20) but had no effect on Jm----s. Both fluxes were linearly related to HCO3- concentration on the buffered side, but the slope for Js----m was 3.5 times that for Jm----s. When tissues were initially set up in HEPES buffer rather than HCO3-, Js----m was 0.12 +/- 0.05 mumol.cm-2.h-1 (n = 6), which is not significantly different from zero. Replacement of Na+ by choline reduced Js----m by 40 +/- 3% (n = 11) and ouabain (1 mM) by 24 +/- 3% (n = 5). Replacement of Cl- with isethionate or K+ with Na+ for 60 min did not alter Js----m. Serosal application of DIDS (0.5 mM) reduced Js----m by 24 +/- 6% (n = 6), but SITS (0.5 mM), furosemide (1 mM), acetazolamide (0.1 mM), amiloride (1 mM), and a proton pump inhibitor (Sch 28080, 50 microM) had no effect. Mucosal application of DIDS, furosemide, and amiloride had no effect on Js----m. Serosal tetrodotoxin (1 microM) and indomethacin (28 microM) were also without effect.(ABSTRACT TRUNCATED AT 250 WORDS)