<0.001), lysozyme (p=0.006), and IgA (p<0.001) than those of cow's milk formula-fed infants. The concentrations of total and secretory IgA were correlated significantly (r=0.88, p<0.001) and 95% of total IgA was SIgA. The fecal concentration of specific SIgA antibodies to E. coli O antigens in FM-fed infants correlated with the concentration of these antibodies in their milk (p<0.001). However, there were no direct relationships between the milk concentrations or the infants' intakes of the other selected immune factors and the excretion of these factors in the feces. Significant relationships were noted among the immune factors in the feces, but not among the concentrations of these factors in the fortified human milk. The increased quantity of selected immune factors in the feces of very low birth weight infants fed FM may have resulted not only from passive ingestion and persistence of these factors throughout the gastrointestinal tract but also from endogenous synthesis induced by the feeding of human milk.