To determine the optimal method for tube-feeding premature infants we conducted a randomized trial of bolus (B) vs continuous (C) feeding in premature infants stratified by gestational age and intent to breast feed. Once feeding was initiated, infants maintained their respective feeding method until full oral feeding was established. Growth, feeding tolerance, medication usage, nutrient absorption, lactase activity (lactose/lactulose ratio in urine), and intestinal permeability (mannitol/lactulose ratio in urine) were assessed. Groups B (n=81) and C (n=77) were similar in birth weight (1.0± 0.2 kg, mean ± SD), gestation (28 ± 1 wk), receipt of antenatal steroids (B:n=39,C:n=47), deaths (B:n=6,C:n=4), necrotizing enterocolitis (B:n=13,C:n=8), sepsis (B:n=31,C:n=36), need for GI surgery(B:n=9,C:n=10), and receipt of oxygen, steroids, diuretics, and duration of parenteral nutrition. Milk was initiated at 11 ± 6 d in each group. Group B achieved full tube-feeding sooner than C (26 ± 8 vs 35 ± 29 d, p=0.05). Feeding tolerance, as measured by the% occurrence of any gastric residual volume > 0 ml (41 ± 12% vs 56 ± 12%,p<0.001), > 2 ml/kg (9 ± 8% vs 17 ± 7%,p<0.001), and > 50% of 3-h feeding (1 ± 2% vs 2± 3%, p=0.04) was significantly more common in B than C. Group B needed significantly less therapy for gastroesophageal reflux than C(p=0.03). There were no differences in the absorption and retention of nitrogen, fat, calcium, phosphorus, zinc, or copper between Groups B and C. Urinary excretion of Ca, however, was significantly greater in B vs C (4± 3 vs 3 ± 2 and 7 ± 5 vs 4 ± 2 mg/kg/d, at 6 and 9 wk, respectively, p<0.01). Lactase activity, intestinal permeability, and growth were not affected by feeding method. These data indicate that the bolus feeding method is associated with more rapid attainment of full tube-feeding, better feeding tolerance, less need for anti-reflux therapy but no advantage with respect to growth.