PIP: Some studies have shown that premature, low birth weight (LBW) babies fed with pooled banked human milk had a significantly lower growth than those fed with commercial formula. Such studies have often utilized donated milk, but not milk from mothers of premature babies. This study compares the composition of milk of mothers who have delivered premature infants with milk donated to a milk bank. Milk was collected from 16 mothers of premature infants at each feeding and pooled and frozen in batches corresponding to the following postpartum ages: 1-3, 4-6, 7-10, 11-13, 14-16, 17-21, 22-28, 29-42, 43-56, and 57-98 days. Milk donated from mothers of term infants was also collected at each feeding and pooled, without attempt to group it into various postpartum ages. Total nitrogen, nonprotein nitrogen, sodium, potassium, and total calcium quantities were determined. Total nitrogen content in the milk of mothers of premature babies was significantly higher during the 1st 2 weeks of lactation, and regressed toward milk donor values by 1 month. Nonprotein nitrogen content of premature infants' mothers' milk was significantly higher during the 1st 3 days of lactation than the donated pool milk. Sodium content was rather elevated for the 1st 3 days and lower after 2 months than the donated pooled breast milk. Potassium contents were similar, and calcium content was higher in the milk of mothers of premature infants for the 1st 3 months. To insure a higher intake of total nitrogen in premature babies, the excess milk supply during the 1st 2 weeks of lactation, i.e. when it is richer in total nitrogen, could be appropriately stored and consequently fed to the premature infant to provide positive nitrogen balance. Therefore, the earlier calculations about premature infants' growth will have to be recalculated if this method is used. It remains to be investigated whether the lower content of nonprotein nitrogen in the milk of premature infants' mothers will influence the growth process.