The aim of this study was to compare the effects of pirenzepine with those of atropine a non-selective antimuscarinic agent, on gastroduodenal motor patterns in duodenal ulcer patients. Twenty patients were allocated at random to 2 groups of 10 subjects each. The drugs were administered by bolus intravenous injection as equiactive antisecretory doses of 10 mg pirenzepine and 1 mg atropine. Before and 15 min after drug administration all patients underwent a gastroduodenal manometric and reflexogenic study with a specially designed probe and three inflatable latex balloons. Both drugs significantly decreased antral and duodenal pressure, but atropine was much more effective than pirenzepine: 91 +/- 2% verus 54 +/- 9% decrease in the motility index for the antrum and 95 +/- 1% versus 49 +/- 7% for the duodenum (p less than 0.01). The antral motor threshold was not modified by either drug. The results of this study confirm the selectivity of action of pirenzepine on gastric function.