1. A simple method is given for testing the efficiency of a salt fractionation method for the separation of albumin from globulins in serum or plasma. This consists of electrophoresis of the concentrated filtrate obtained by treatment of the specimen with the fractionating agent. 2. A method is presented for total precipitation of globulins in serum, or plasma with the use of a 27.2% sodium sulfate solution. Electrophoresis of the concentrated filtrates from such preparations reveals similarities close to the electrophoresis of pure albumin. 3. A comparison is made between the proposed salt fractionation technic and the results of electrophoresis in a series of 25 specimens. In this comparison, allowance of albumin trailing during protein paper electrophoresis is made. It is thought that the degree of agreement between the two methods would not have been so favorable if such an allowance had not been made.