To assess the effect of duodenal ulcer surgery on Campylobacter pyloridis gastric biopsies were done and fasting bile acid concentrations in gastric aspirates were measured in 35 patients with active duodenal ulceration and 54 who had undergone surgery at some time. Biopsy specimens were assessed blind for the presence of C pyloridis and scored for severity of reflux gastritis by the use of a histological grading system. Among patients who had undergone highly selective vagotomy the proportion who were C pyloridis-positive was similar to that in the unoperated group, but among those who had undergone Billroth I partial gastrectomy, Billroth II partial gastrectomy, or truncal vagotomy and gastroenterostomy it was significantly lower (p less than 0.001). The absence of C pyloridis correlated strongly (p less than 0.001) with high reflux scores and increased bile acid concentrations in the stomach. Reflux scores and bile acid concentrations were significantly higher (p less than 0.01) after Billroth I and Billroth II partial gastrectomies and truncal vagotomy and gastroenterostomy than in the active duodenal ulcer or highly selective vagotomy groups. There was a highly significant correlation (p less than 0.001) between reflux scores and bile acid concentrations. These results suggest that reflux may disrupt mucus and thus cause the death of campylobacters that live beneath it. They also suggest that reflux may produce a reflux-specific gastritis. Highly selective vagotomy may protect against these changes in the gastric mucosa.