Routine transpyloric placement of feeding tubes reduces aspiration in intensive care unit patients. Spontaneous passage eliminates the need for radiologic or endoscopic intervention. It is unclear whether the addition of a weight to the end of the tube or the use of the prokinetic agent metoclopramide in the conventional dose (10 mg) improves spontaneous transpyloric placement. In a randomized, prospective trial, 39 intensive care unit patients (age > 2 years) had a total of 50 nasoenteral tubes placed after intravenous metoclopramide (20 mg in adults, 0.2 mg/kg in children). The tubes were 8 French in diameter with either a weighted end or an unweighted bullet tip. Tip position was confirmed radiographically within 4 hours after blinded placement and at 1 and 2 days if spontaneous passage had not occurred. At 4 hours, 36% of the weighted tubes and 84% of the unweighted tubes (p < .002) had passed through the pylorus. At 1 day, 48% of the weighted tubes and 86% of the unweighted tubes (p < .007) had achieved transpyloric position. At 2 days, 56% and 92% of the weighted and unweighted tubes, respectively, had passed through the pylorus (p < .009, chi 2). The occurrence of transpyloric passage and the rapidity at which it occurred was significantly greater for the unweighted tube during all time periods. A poststudy trial was conducted to evaluate the effect of the normal dose of metoclopramide (10 mg in adults and 0.1 mg/kg in children) and the transpyloric passage of the unweighted feeding tube. Twenty-five patients received 10 mg of metoclopramide before the insertion of the unweighted tube.(ABSTRACT TRUNCATED AT 250 WORDS)