BACKGROUND We tested whether direct transcutaneous bilirubin (TcB) measurement from an area unexposed to phototherapy is reliable for estimation of total serum bilirubin (TSB) in neonates during phototherapy and whether it contributes to reduction in TSB blood sampling in phototherapy decision making. METHODS This was a retrospective observational study of term neonates who received phototherapy in the mother's room. TSB and TcB from the neonate's sternum were measured before and during phototherapy and compared using linear regression analysis and Bland-Altman plot, respectively. Various cut-offs of TcB for estimating TSB during phototherapy at >72 h after birth were analyzed. RESULTS There were moderate correlations between TSB and TcB before (r = 0.56) and during (r = 0.47) phototherapy in 125 neonates. The mean difference (TSB-TcB) before and during phototherapy was 1.2 ± 1.7 mg/dL and 1.0 ± 1.7 mg/dL, respectively. The 95% limits of agreement for the difference before and during phototherapy ranged from -2.1 to 4.5 and from -2.3 to 4.3 mg/dL, respectively. For TSB ≤18 mg/dL during phototherapy, a TcB cut-off of 14 mg/dL had a specificity of 1.0; with this method, 43% of the TSB measurements could have been avoided. CONCLUSIONS Direct measurement of TcB during phototherapy using a bed-type device is a reliable method to estimate TSB in term neonates and would contribute to a reduction in blood sampling. It cannot, however, be used as a substitute for TSB measurement.