The study aim was to evaluate urinary excretion of Retinol Binding Protein (RBP), compared with urinary excretion of N-acetyl-beta-D-glucosaminidase (NAG), in preterm infants with anoxia and netilmicin treatment. Urinary RBP and NAG were evaluated in 83 preterm newborns divided in 4 groups: 37 healthy preterm newborns (controls); 14 with neonatal anoxia; 16 treated with ampicillin + netilmicin; 16 with neonatal anoxia and treated with ampicillin + netilmicin. RBP was determined by an automated nephelometric technique and NAG by a colorimetric method on 5-h urine samples in the first week of life. Results showed that urinary excretion of RBP (average from first week values) was 1.06+/-0.67 g/mol creatinine (mean +/- SD) in controls, 1.99+/-1.41 in antibiotic-treated newborns, 3.99+/-4.57 in anoxic newborns and 3.75+/-3.48 in anoxic newborns under antibiotic treatment. When gestational age was not considered, a marked effect of anoxia (P<0.001) and a borderline effect of netilmicin (P<0.059) on RBP excretion were detected by ANOVA. However when gestational age was also considered by analysis of covariance, it appeared as the strongest predictor of RBP excretion (P<0.001), while the effect of netilmicin was no longer significant (P=0.181). The effect of anoxia persisted, although less remarkable (P=0.010). Conversely anoxia did not affect urinary NAG excretion, which was rather correlated with gestational age and netilmicin administration. The authors conclude that RBP and NAG urinary excretion may be used to discriminate between neonatal anoxia and netilmicin treatment, respectively as etiologic factors of renal tubular damage in the newborn.