I investigated the immediate ( less than 30s) reaction between bromcresol green and serum, as a measure of albumin content. Fifty sera were analyzed in duplicate for albumin by electrophoresis on cellulose acetate by a procedure previously shown to give results in good agreement with two immunoprecipitation techniques, and by a bromcresol green method with the absorbance being read as soon as possible after mixing. Results by the dye method were 3.0 g/liter higher than those by electrophoresis; this difference was independent of the albumin content of the serum. Slight turbidity, jaundice, or added hemoglobin did not significantly interfere with albumin determination. Aparently the immediate reaction between serum and the dye reagent provides a simple, reliable measure of albumin content when 3.0 g/liter is subtracted from the result obtained.