Although nutrient balances are used to determine nutrient assimilation, the balance technique is not well described in the preterm infant. Our purpose is to describe the balance technique; discuss factors that may interfere with complete balance collections; and determine the effects of intake, patient characteristics, and formula fed on nutrient assimilation in the preterm infant. Fifty-four 84-h balances were performed in a group of well premature infants receiving either a casein- (Similac with Iron) or whey-predominant (Similac with Whey plus Iron) infant formula. Mean (+/- SD; range) birth weight, gestation, postnatal age, and balance weight were 1,320 g (+/- 200; 820-1,720), 32.0 weeks (+/- 2.0; 27-34), 36 days (+/- 14.8; 15-87), and 1,729 g (+/- 146; 1,435-2,070), respectively. Although absorption and retention varied, results were comparable with previous studies in the preterm infant. Results of linear regression analysis, using the stepwise regression procedure, indicate that variability in nutrient intake (mg/kg/day) accounted for 97 and 84% of the variability in nitrogen absorption and retention (mg/kg/day). Variability in nutrient intake (mg/kg/day) also accounted for 96 and 57% of the variability in phosphorus absorption and retention (mg/kg/day). Effects due to intake, birth weight, gestation, postnatal age, balance weight, and formula fed also accounted for a substantial part of the variability in calcium, zinc, and copper absorption and retention. The results of this study indicate that factors other than birth weight, gestation, and the type of formula fed have a significant effect on absorption and retention of nutrients and should be considered as confounding variables when interpreting effects due to treatment.