We used the rabbit esophagus as a model to study clinically significant variables which may be important in the pathophysiology of reflux esophagitis in man.* Increased permeability to hydrogen ion, as indicated by a net acid flux (NAF) out of the lumen, was our index of injury. NAF was measured with an in-line pH stat autoburette in rabbits lightly anesthetized with Innovar. We found minimal NAF at pH 5 and 2, but NAF was 70±11 μEq/10 min at pH 1. At pH 2, the addition of bile salts resulted in the appearance of a significant increase in NAF. When we varied the concentration of bile salts from 0 to 5 mM, we found that NAF (μEq/10 min) varied directly and significantly with the bile salt concentration (B, mM) according to this relationship: NAF=12.7 (B)−0.8 (r=0.96,P<0.01). When we varied the duration of time (t, min) for which the mucosa was exposed to bile salts, we found NAF varied directly and significantly with time from 1 to 60 min according to the following relationship: NAF=1.25t+10.8 (r=0.93,P<0.05). In other experiments, we demonstrated that the taurine-conjugated bile salts taurodeoxycholate and taurocholate increased NAF at pH 2, while the unconjugated forms did not. The unconjugated forms did increase NAF at pH 7. This difference is attributable to the lower pKa of the taurine-conjugated bile salts. Deoxycholate increased NAF more than cholate. In none of our experiments did we see evidence that the esophageal mucosa could recover from bile salt-induced injury.