In the course of investigations on the chemistry of the liver proteins we have had occasion to determine the liver-albumin content under a variety of conditions. The fundamental problem which we desired to elucidate was whether liver albumin exists as a preformed protein or whether it arises as an artefact during fractionation. It occurred to us that a partial answer would be obtained by studying the ratio, albumin/total salt-soluble protein, at different hydrogen ion concentrations maintained during the preliminary sodium chloride extractions. It is known that the total salt-soluble protein extractable from liver increases greatly with increase of pH over the range pH 4 to 8. If the liver albumin were largely an artefact, arising as a dissociation product from the salt-soluble protein fraction, one might reasonably expect that increases in the latter would be reflected by equi-proportional increases in the former; the ratio, albumin/total salt-soluble protein, would remain constant. Although the present work does not give a decisive answer to the question it yields certain information about the liver-albumin content which is pertinent and significant. For the purposes of this investigation we used dog liver, rapidly perfused in situ to remove blood, then excised, frozen with liquid air, powdered, and preserved at −10°C. In the first group of experiments the procedure followed was essentially that reported previously, except that 3 successive extractions with sodium chloride were employed instead of 2, and the period of dialysis was shortened to 2 days. Extractions were made at pH 4.3, 4.7, 5.0, 6.4, and 7.0. Between pH 4.7 and 7.0 the albumin content remained virtually constant at about 2.2% (extremes 2.07, 2.33); the total salt-soluble protein content increased progressively from 6.44 to 9.18%; the ratio, albumin/total salt-soluble protein, decreased from 0.36 to 0.23.