While abnormalities in antroduodenal motor function have been documented in both organic and "functional" disorders, controversy surrounds the ideal manometric technique. We sought, therefore, to evaluate a digital solid-state ambulatory system. Sixteen normal volunteers underwent 24-hr recordings of antroduodenal motility. Following catheter placement, a standardized meal was ingested in the laboratory; thereafter, subjects were ambulatory and assumed normal diet and activities. The system was well tolerated; subjects reported that it did not affect their usual activities. Migrating motor complex (MMC) activity was identified in each subject (mean frequency: 4.1 MMCs/24 hr, range 1-8); on average 1.9 (range 0-4, frequency 0.1/hr) occurred while awake and 2.1 (range 0-5, 0.3/hr, P < 0.05 vs awake) during sleep. The fed response was evaluated by calculating a motility index (MI) at 30-min intervals from 30 min before to 120 min following meal ingestion. Postprandially, MI was maximal during the first 30 min following meal ingestion: MI (mean +/- SD) 30 min before vs 30 min after meal in the antrum: 4.16 +/- 1.42 vs 5.33 +/- 0.72 (P < 0.05), duodenum: 4.04 +/- 0.80 vs 4.57 +/- 0.47 (P < 0.05), respectively. None of the other postprandial intervals were significantly different from baseline. There was no significant difference in MI between the standard and ad libitum meals. Retrograde catheter migration (mean 5.6, range 1-10 cm) occurred in relation to all meals: as a consequence, antral recordings were lost following 60% of all meals, thereby limiting meaningful analysis of the antral fed response. We conclude, firstly, that while an ambulatory antroduodenal manometry system is well tolerated and reliably records duodenal motility, postprandial catheter migration limits antral recordings, and, secondly, that a motility index calculated during the first 30 min following an ad libitum meal accurately reflects the fed motor response.