In this study, an effort has been made to predict the solid-liquid equilibrium (SLE) behavior of different solids (pharmaceuticals) in many common solvents and their mixtures. A modified optimization of a recent thermodynamic model, the NRTL-SAC model, was used in all stages of calculation (VLE, LLE, and SLE predictions). The batch cooling-antisolvent crystallization process was simulated for seven model molecules from the initial temperature to the final temperature and for the volume fraction of each solvent. The feasible region of temperature for each crystallization case was calculated based on the bubble-point temperature of the solvent mixture and the melting point of the model molecules. The NRTL-SAC model was used in conjunction with the optimization procedure to test the complete miscibility of solvents during each part of crystallization. After estimating the optimum solvent mixture (combination) for a specific model molecule, the results for single, binary, and ternary solvent mixtures were compared. The results obtained from the binary and ternary combinations were similar in terms of crystallization yields per mass of solvent mixture and far superior to those obtained with single solvents. The proposed algorithm demonstrates flexibility, simplicity, and accuracy in predicting the phase behavior and eventual optimal solvent screening for the crystallization of pharmaceutical components.