Epidemiological evidence has suggested that the declining prevalence of duodenal ulcer disease may be attributable to rising consumption of polyunsaturated fatty acids, a hypothesis supported by in vitro evidence of toxicity of such substances to Helicobacter pylori. The objective of the present study was to establish whether this association is causal. Forty patients with proven infection with H. pylori and endoscopic evidence of past or present duodenal ulcer disease were randomized to receive either polyunsaturated fatty acids (PUFA group), in the form of capsules and margarine, or a placebo (control). Both groups received concurrent H2 antagonist therapy. Efficacy of therapy was determined endoscopically by assessment of ulcer healing while H. pylori status was determined by antral biopsy, urease (EC 188.8.131.52) culture and histological assessment of the severity of H. pylori infection. Antral levels of prostaglandin E2 (PGE2) and leukotriene B4 (LTB4) were quantified. Compliance was monitored. Before treatment, both groups were comparable for severity of H. pylori infection, smoking status and levels of LTB4 and PGE2. Despite a significant difference in consumption of linoleic acid (19.9 (se) 1.6) g for PUFA group ν. 6.7 (se 0.8) g for controls (P < 0.01) and linolenic acid (2.6 (se) 0.2) g ν. 0.6 (se 0.03) g (P < 0.01) there was no significant change in either the severity of H. pylori infection or prostaglandin levels in either group at 6 weeks. Consumption of a considerable amount of PUFA does not inhibit the colonization of the stomach by H. pylori nor does this alter the inflammatory changes characteristic of H. pylori gastritis. We conclude that the association between duodenal ulceration and a low level of dietary PUFA is likely to be spurious, probably reflecting the effect of confounding factors such as affluence, social class or smoking.