The limulus amebocyte lysate test for endotoxin was done on samples of blood from 22 well babies and 33 neonates in an intensive-care nursery. The objective was to determine whether falsely positive test results occurred in samples from newborn infants during acquisition of usual bowel flora. One neonate had a transiently positive limulus amebocyte lysate test; he had clinical signs of spesis, but no bacteremia could be documented. Unlike the nitroblue tetrazolium test, the limulus amebocyte lysate test does not appear to be regularly positive in neonates. The sensitivity of the test in detection of endotoxemia associated with gram-negative spesis in the neonate remains to be determined in a large prospective study.