The outcome of infants with TPN was assessed in a case-control study. The study group (SG) included 116 patients with TPN and the control group (CG) 223: patients were treated in the neonatal ICU at least for 48 h without TPN, paired by birthweight and gestational age. SG presented a high incidence(P<0.001) of MD in (97/116 vs 104/223), hyperglycemia and hypocalcemia were more frequent. No differences were found in incidence of sepsis between both groups, but SG had more Candida-positive blood cultures: 8% vs 1% in CG(p<0.0002) and S. epidermidis: 14% vs CG 5% (p<0.0004). IPPB was used in TPN: 84% vs. CG: 42% (p<0.0001). Mortality was higher in CG during the first week: 5% in SG vs. 15% in CG (p<0.01), but later increased in SG (30% vs. 9%, p<0.0001). This study showed that TPN in high risk newborns is safe but requires careful monitoring. The increased mortality and the need of respiratory support in the SG was related with other illnesses present in the patients.