OBJECTIVE The aim of this prospective study was to evaluate the accuracy of transcutaneous bilirubinometry using the Minolta Air-Shields JM-103 device in preterm newborns of gestational age 32-34 weeks, and to identify the most appropriate measurement site. METHODS Transcutaneous bilirubin (TcB) measurements were performed over forehead, sternum and abdomen, if total serum bilirubin (TSB) had to be determined on clinical indication in neonates of selected gestational age. TSB levels were measured in a clinical laboratory using direct spectrophotometry. In order to assess transcutaneous bilirubinometry accuracy, differences between TSB and TcB, their CI 95%, and correlation coefficients (r) between TcB and TSB were evaluated. RESULTS The study group consisted of 44 infants, including 6 very low birth weight (VLBW) neonates. The correlations between transcutaneous and laboratory values were found to be significant and close. Minimal differences were observed when measured over sternum. The measurements over forehead had a tendency to underestimate TSB levels. CONCLUSIONS Noninvasive measurement by Minolta JM-103 demonstrated significant accuracy. The authors recommend measurements over sternum or abdomen in premature infants born within 32-34 gestational weeks as a reliable and accurate neonatal hyperbilirubinemia screening test. Transcutaneous bilirubinometry has the potential to reduce the number of blood samplings, thus reducing neonatal pain and discomfort, parental distress and medical care cost.