Entrapment efficiency in protein microencapsulation into biodegradable polymer microspheres depends on the nature of the solvents used for dissolving the polymer. In such a system, three main interactions are present, i.e., polymer-water (aqueous protein solution), polymer-solvent, and solvent-water (aqueous protein solution). These interactions are quantified by the three interaction energies, Delta(int)E(p), Delta(int)E(1), and Delta(int)E(2), respectively. For a given polymer, Delta(int)E(p) is constant, but Delta(int)E(1) and Delta(int)E(2) vary as a function of the nature of the solvent used. In this contribution, the model protein bovine serum albumin was microencapsulated into poly(D,L-lactic acid) by spray-drying. It was demonstrated that increasing absolute values of the sum Delta(int)E(1) + Delta(int)E(2) leads to decreasing encapsulation efficiencies. Delta(int)E(2) was estimated from the heat of dissolution of the polymer in the selected solvents and the energy of cavity formation, and Delta(int)E(2) from delta(d) and delta(p) and Drago’s parameters E and C. A linear fit between the sum of interaction energies, equivalent to Delta(int)E(1) + Delta(int)E(2), and the actual microencapsulation efficiency gave a reasonable correlation with a correlation coefficient r = 0.954. This represents an acceptable correlation considering enthalpies were used to predict interaction energies. Moreover, if microencapsulation efficiency is correlated directly with Delta(int)E(1) (expressed by its equivalent calculated value), and the parameters delta(d), delta(p), E, and C, instead of Delta(int)E(2), an even better correlation with a multiple r = 0.9995 is observed. This rational approach in microencapsulation is of high importance as it is based on thermodynamic parameters.