Urinary lithogenic and inhibitory factors were studied in 27 preterm infants; 16 had total parenteral nutrition (TPN) and 11 had breastmilk with an additional glucose-sodium chloride infusion. Urines were collected for 24 hours on day 2 (period A), day 3 (B), and once between days 4 and 10 (C). Urinary calcium oxalate saturation was calculated by the computer program EQUIL 2. Renal ultrasonography was performed every second week until discharge. The calcium/creatinine ratio increased in infants on TPN (A 0.91; C 1.68 mol/mol) and was significantly higher at period C than that in infants on breastmilk/infusion (A 0.52; C 0.36). The oxalate/creatinine ratio was persistently higher with TPN (203 mmol/mol) than with breastmilk/infusion (98; 137). The citrate/creatinine remained constant with TPN (0.44 mol/mol), whereas it increased significantly with breastmilk/infusion (0.26; 0.49). Calcium/citrate rose considerably with TPN, but decreased with breastmilk/infusion to a significantly lower level than with TPN. The urinary calcium oxalate saturation increased with TPN (2.4; 4.5) and decreased with breastmilk/infusion (2.1; 1.5) to a significantly lower value than with TPN. Nephrocalcinosis developed in two infants on TPN. Mean daily calcium intake was similar in both groups, whereas protein, sodium, and phosphorus intake were significantly higher on TPN. It is concluded that the increase in urinary calcium oxalate saturation observed with TPN is due to the combined effect of an increased urinary calcium excretion and higher urinary oxalate/creatinine and calcium/citrate ratios. The changes observed are likely to be caused by TPN itself, which differs in several respects from breastmilk feeding.