To compare the performance of Modified Kramer's and Kramer's methods in terms of agreement with total serum bilirubin (TSB). This cross-sectional study was done in Level-III neonatal unit in New Delhi. Visibly jaundiced neonates born at ≥35 wk of gestation were enrolled and examined by (i) conventional Kramer's, (ii) Modified Kramer's in artificial (MK-A) and (iii) natural daylight (MK-N), and finally sampled for estimation of TSB by point-of-care spectrophotometry. The primary outcomes were agreement of Kramer's and Modified Kramer's with TSB and accuracy in terms of proportion of bilirubin estimates lying within ±2mg/dL of TSB; secondary outcome was agreement of MK-A and MK-N with TSB. A total of 144 neonates with median gestation of 37 wk and mean birth weight of 2788g were enrolled. Bland Altman analysis between Kramer's and TSB yielded mean difference of 1.7mg/dL, 95% limits of agreement (LOA) -3.1 to 6.6mg/dL. For Modified Kramer's and TSB, mean difference was -0.02mg/dL, 95% LOA -4.7 to 4.7mg/dL under artificial light; 0.02mg/dL, 95% LOA -4.2 to 4.2mg/dL under natural daylight. MK-N had highest proportion of bilirubin estimates lying within ±2mg/dL of TSB (68.7%) as compared to MK-A (59.7%) [OR, 1.77; 95% CI, 1.09 to 2.86] and Kramer's (45.8%) [OR, 1.65; 95% CI, 1.27 to 2.15]. Though all the three methods had poor agreement with TSB, Modified Kramer's method when performed in natural light had reasonable accuracy, however limited clinical utility, in evaluation of clinical jaundice.