Interactions mediated by layers of -casein and para--casein have been calculated and compared using the self consistent field (SCF) approach, at pH and electrolyte concentrations found in milk. Our results show the interaction potentials at close separation distances to be repulsive for -casein, while they are attractive for para--casein. We have also studied -casein chains stripped of their sugar moieties and highlighted the important role that the carbohydrate side chains play in the provision of steric forces. The point of transition from a repulsive to an attractive interaction is found to be rather sensitive to the degree of coverage of the surface by the protein chains. At =0.0025 chains per unit monomer area, the value is just over 40%, whereas at =0.0032 this increases to 87%. At the coverage values higher than those we estimate for the surface of casein micelles, no transition is detected and the interactions remain repulsive even when all of -casein has been converted to para--casein, as already shown by Mellema, Leermakers, & de Kruif (1999). At lower surface coverage values, the interaction potential curves for otherwise uncharged surfaces, covered with para--casein, posses an energy barrier. We have demonstrated that the origin of this energy barrier is electrostatic, with para--casein contributing a net positive charge to the surface.We have also explored the importance of the asymmetry in the distribution of charge between the para--casein and the glycomacropeptide sides of the protein. The glycomacropeptide is net negative while para--casein side is net positive at neutral pH. Our results suggest that on a negatively charged surface, such as the surface of casein micelles, this uneven distribution of charge is just as important in determining the conformation of -casein chains as is the difference in the hydrophobic and hydrophilic nature of the two sides of this protein.