Dipolar interactions significantly influence the phase behavior of many systems of interest to the biochemical, chemical, petroleum and polymer industries. For example, the solution behavior of amino acids, small peptides, polar solvents and co-polymers have potential applications for biochemicals, water-soluble polymers for paints and coatings, and surfactants. By considering molecular-level interactions, the phase behavior of a large range of systems can be predicted. Using structural analysis and thermodynamics, the essential role of polar functional groups on solubility of small biochemicals is established. An accurate model for fluid mixtures with multiple polar functional groups is developed. Results from a systematic experimental study on the aqueous solubility of amino acids and dipeptides as a function of temperature, salt type, and salt concentration are analyzed. Changes in temperature and residue sequence have the most substantial effect on solubility. Structural analysis shows that intra and intermolecular association largely influence the behavior. For small molecules, end effects dominate the behavior but should be less important for many biochemicals. Values for enthalpic changes from the solid to the infinitely dilute liquid state for the dipeptides of asp and gly are reported. An accurate model for mixtures of polar fluids, in which any number of groups and/or any component of the mixture may be polar, is developed by applying the u-expansion to a reference fluid mixture of polar and non-polar spheres. An in-depth parameter study of this model, named Polar SAFT, for a homologous series of ketones indicates that the model parameters have physically reasonable values. A methodology is proposed for developing a group-contribution approach for the model. The ability of Polar SAFT to accurately predict the effect of multiple dipolar groups and molecular shape on the phase behavior of binary mixtures of polar and non-polar components is exemplified by application to a series of ketone/alkane mixtures. Using only pure component parameters, Polar SAFT accurately represents these systems, indicating the predictive capability of the model and the importance of explicitly accounting for polar interactions. For alkane/copolymer (poly(ethylene-co-methyl acrylate)) solutions, Polar SAFT accurately predicts the polar co-monomer content and solvent effects on cloud point behavior.