The existence of preduodenal phases of gallbladder emptying in normal volunteers was investigated by evaluating the effect of sham feeding a sandwich (chew and spit) (n = 15), the sight and smell of food followed by sham feeding a cooked meal (n = 15), and gastric distension (intragastric balloon tube) (n = 9) on gallbladder emptying and comparing these responses with those after ingestion of a standard meal (n = 14). A control group given no intestinal stimuli were studied to determine the frequency of spontaneous emptying during fasting (n = 18). 99mTc-EHIDA (2,6 diethylphenylcarbamoylmethyliminodiacetic acid) was used as the biliary tracer. The frequency of gallbladder emptying during fasting was 0.0045/min. Thus, in any 20 min period emptying occurred spontaneously in only about 1 in 11 volunteers. Significant emptying (greater than 5 per cent over 20 min) occurred in 8 out of 15 volunteers after sham feeding a sandwich (P less than 0.001 versus control), in 8 out of 15 volunteers after sham feeding a cooked meal (P = 1.0 versus sandwich sham feed, n.s.), in 6 out of 9 volunteers after gastric distension (P less than 0.001 versus control) and in 14 out of 14 volunteers after meal ingestion. The rate of emptying was significantly greater in the gastric distension and meal ingestion groups compared with other groups (P less than 0.05). However, the emptying rate after sham feeding was not significantly different from spontaneous gallbladder emptying (P greater than 0.05, n.s.). There was no significant difference in time to onset of emptying between the four stimulus groups (P greater than 0.05, n.s.). These results suggest that pre-duodenal stimuli can evoke gallbladder emptying in man, although the precise physiological significance of this phenomenon remains to be established.