Bush and Reed reported (Clin Chem 1987;33:821-3) that the reaction of albumin with bromcresol purple but not with bromcresol green underestimated the concentration of albumin in synthetically obtained bilirubin-albumin (Bd) by 29%. Their unproven assumption was that chemically synthesized Bd behaved in a manner indistinguishable from the natural Bd in icteric serum. Here we verify that Bd, whether synthetically obtained or isolated from serum, causes an underestimation of albumin in the bromcresol purple but not in the bromcresol green method. The molar ratio of Bd from either source to underestimated albumin approximates 1.0, suggesting that one molecule of Bd would react equivalently to a molecule of albumin in the bromcresol purple method. This underestimation might falsely suggest hypoalbuminemia in patients with increased serum Bd.