Dear editor, the article by Keren et al 'Visual assessment of jaundice in term and late preterm infants' justifies the limitations of visual assessment of jaundice in newborns. The authors have advised to do serum bilirubin levels in case of visible jaundice. The problem with total serum bilirubin level is that it is dependent on the serum albumin level. Therefore a healthy newborn with a good serum albumin will bind bilirubin which will appear in the circulation showing a high serum bilirubin level despite having low levels in the tissues as compared to a preterm or sick neonate with low albumin levels where despite low serum bilirubin level due to poor binding to albumin and more tissue bound bilirubin, there is an increased risk of damage to the tissues including the brain. Also, the laboratory estimations of bilirubin are quite variable. Till free bilirubin measurements are available we are forced to use these surrogate markers of bilirubin which are far from being perfect in predicting the brain damage.