Parameters determined from binary experiments were used to predict the behavior of multicomponent A/B/A-C polymer blends, where A is saturated polybutadiene with 90% 1,2-addition (sPB90), B is polyisobutylene (PIB), and C is also saturated polybutadiene but with 63% 1,2-addition (sPB63). The polymers were chosen such that the binary interactions (A/B, A/C, and B/C) are analogous to those in oil (A)/water (B)/nonionic surfactant (A-C) systems, where A/B and A/C are unfavorable interactions (chi > 0) and B/C is a favorable interaction (chi < 0). The Flory-Huggins interaction parameters (chi(AB), chi(AG), and chi(BC)) and the statistical segment lengths (l(A), l(B), and l(C)) were all determined experimentally by fitting the random phase approximation (RPA) to small-angle neutron scattering (SANS) data from the three binary homopolymer blends. These parameters were successfully used to predict the scattering from concentration fluctuations in a homogeneous A/B/A-C blend using multicomponent RPA. These same binary parameters were also used as the only inputs to self-consistent field theory (SCFT) calculations of ordered multicomponent polymer blends. The SCFT calculations enabled quantitative interpretation of the SANS profiles from microphase separated A/B/A-C blends. The phase separation temperatures predicted by theory for the blends were within the experimental error, and the theoretical domain spacings were within 10% of the experimental values.