Unbound bilirubin(UB) was measured in the blood of 79 infants using a micro automated modified peroxidase method employing the UB Analyzer developed by Nakamura.This simple micromethod was found to correlate with the manual method and showed very good reproducibility(Coeff.var.4%). Several factors influencing the level of unbound bilirubin were examined,including gestational age,birth wt.,sick vs. well,post-natal age,total bilirubin(TB), total protein(TP) ,and phototherapy.Direct correlation was found between TB and UB (r=0.69,p=.0001,n(sample)=145)and negative correlation with total protein vs UB(r=-0.24,p=0.001,n=l52).The ratio TB/TP vs UB correlated directly(r=0.88,p=0.0001,n=145).Unbound bilirubin was higher in the first 168 hrs of life than in infants>168 hrs of life.At the same total bilirubin level,there was a dramatic difference in the amount of unbound bilirubin in blood of infants of BW<1000 g than in infants≥1000 g.There was a sequential decrease in UB with increasing BW. Sick infants ,had higher UB than did well infants in the same weight group. At a given serum bilirubin level, unbound bilirubin was higher before phototherapy than during or after phototherapy.It is clear that several factors in addition to total bilirubin determine the concentration of unbound bilirubin in the blood of newborn infants. The UB Analyzer affords a simple reproducible micromethod for measurement of UB in the clinical setting. The ratio TB/TP corresponded better with UB than did TB alone and could be useful as a screen for UB.