IN the course of studies on the gel filtration of a number of proteins on ‘Sephadex G–50’ (trade name for a cross-linked polysaccharide dextran made by Pharmacia, Uppsala, Sweden, who publish a booklet containing an extensive bibliography), it was found that in the absence of salts some proteins were strongly adsorbed to the gel. For example, ribonuclease, lysozyme, trypsin, bovine serum albumin and a polytyrosyl derivative of trypsin1 were all adsorbed and could not be removed by extensive washing with distilled water. However, these proteins were recovered quantitatively by elution with dilute sodium chloride solutions. Since ‘Sephadex’ has been used extensively for the desalting of protein solutions and for the separation of proteins from materials of low molecular weight, it is important to note that in attempts to remove salts completely by using distilled water as eluant, this phenomenon may lead to complications.