The present study attempts to clarify one of the important issues in the field of metallic glasses (or amorphous alloys), i.e., to predict the glass-forming ability/range (GFA/GFR) of the binary metal systems. Firstly, a brief summary of the experimental observations indicates that in the miscible/immiscible systems, amorphous alloys can be formed in a broad composition range extending from central portion to nearby the edges of the two terminal solid solutions. Consequently, to predict RFA/GFR of a system becomes an issue of determining the critical solid solubility of the system, beyond which a solid solution would collapse into an energetically favored amorphous state. Secondly, the n-body potentials are derived by routine methods for the miscible Ni-Ta, Ni-Mo, and Ni-Ti systems, whereas for the immiscible Ag-Co and Cu-W systems having no any equilibrium alloy, the cross potentials are fitted to some physical properties acquired by first principles calculation for a few possible nonequilibrium alloys. Thirdly, applying the proven realistic potentials, molecular dynamics simulations with solid solution models reveal that the physical origin of crystal-to-amorphous transition is the lattice collapsing while solute atoms exceeding the critical solid solubility and that the GFRs of the systems are within the composition ranges bounded by the two determined critical solid solubilities. It turns out that the predictions are in good agreement with the experimental observations. (C) 2004 Kluwer Academic Publishers.