The heterogeneity of casein is discussed in the light of methods currently used for the fractionation of casein. In particular, the possible heterogeneity of certain preparations of α-casein is considered. This is important because it has been generally considered that α-casein is the protective colloid which is altered when the enzyme, rennin, acts on casein micelles. Recently, Waugh and von Hippel (1956) have suggested that their new component x-casein, and not α-casein, is the protective colloid. These two viewpoints could be reconciled if α-casein samples previously examined contained x-casein as well. In the present work, a study is made of filter paper electrophoresis, micelle-forming properties, and sedimentation of casein fractions. It is shown that x-casein is concentrated with α-casein in fraction A during the alcohol fractionation method of Hipp et al. (1952). On the other hand fraction B contains α-casein essentially free of x-casein. The a-casein obtained in the urea fractionation method of Hipp et al. also contains x-casein. Thus only alcohol fraction B is a suitable source of pure α-casein. During the paper electrophoretic examination of casein fractions a number of minor protein components are observed. A component moving more slowly than γ-casein is present in acid casein, second-cycle casein-fraction P, and an alcohol fraction. This component was first observed in the latter fraction by Hipp et al. (1952) when preparing γ-casein. Electropherograms of second-cycle casein-fraction S indicate the presence of x-, β-, and γ-casein, and two minor components moving between x- and β The way in which these components arise is briefly discussed.