This study was carried out in order to compare the antisecretory effect of a single bedtime dose of roxatidine 150 mg and ranitidine 300 mg and to assess the relationship between the degree and the duration of acid suppression and the healing rates obtained in duodenal ulcer patients treated with the above regimens. Sixty-three patients with endoscopically proven ulcer underwent 24-h gastric pH-metry on day 0, day 1, and day 28 of treatment with both roxatidine and ranitidine. Ulcer healing was checked endoscopically after 4 weeks of therapy. RESULTS: Eight patients did not complete the study, leaving 55 patients eligible for final analysis, 28 in the roxatidine group and 27 in the ranitidine group. Duodenal ulcers were healed in 24--28 (85%) patients of the former and in 22--27 (81%) patients of the latter group (p minus sign NS). Gastric pH was significantly higher (p < 0.001) than basal values on days 1 and 28 with both H2-antagonists. The 24-h pH levels did not differ between day 1 and day 28 with both roxatidine and ranitidine. There was also do difference between the two active treatments. The pattern of gastric acidity significantly differed (p < 0.01) between responder (n = 46) and nonresponder (n = 9) patients to both H2-blockers, and this difference was mainly sustained by nocturnal pH. CONCLUSIONS: A bedtime close of roxatidine 150 mg and ranitidine 300 mg was able to heal more than 80% of duodenal ulcers within 4 weeks of treatment. The lack of tolerance to H2-blockers in duodenal ulcer patients contributes to this good result. The antisecretory effect of H2-antagonists is reduced in nonresponder patients with respect to responder patients and this is mainly due to an impaired control of nocturnal acidity.