The successful development of controlled release formulations for proteins requires that the protein not be denatured during the manufacturing process. The major objective was to develop formulations that stabilize two recombinant human proteins, human growth hormone (rhGH) and interferon-gamma (rhIFN-gamma), at high protein concentrations (> 100 mg/mL) in organic solvents commonly used for microencapsulation, methylene chloride and ethyl acetate. Several excipients were screened to obtain the maximum solubility of each protein. These formulations (aqueous, lyophilized, milled, spray dried, or isoelectric precipitate) were then rapidly screened by emulsification in the organic solvent followed by recovery into excess buffer. Additional screening was performed with solid protein that was suspended in the organic solvent and then recovered with excess buffer. The recovery of native protein was determined by native size exclusion chromatography (SEC-HPLC) and circular dichroism (CD). The selected formulations were encapsulated in polylactic-coglycolic acid (PLGA) microspheres by either water-in-oil-in-water (W/O/W) or solid-in-oil-in-water (S/O/W) methods. The initial protein released from the microspheres incubated at physiological conditions was analyzed by SEC-HPLC, CD, and biological assays. The stability of a given formulation in the rapid screening method correlated well with stability during encapsulation in PLGA microspheres. Formulations of rhGH containing Tween 20 or 80 resulted in lower recovery of native protein, while trehalose and mannitol formulations (phosphate buffer, pH 8.0) yielded complete recovery of native rhGH. Other additives such as carboxymethyl cellulose, gelatin, and dextran 70 were not effective stabilizers, and polyethylene glycol provided some stabilization of rhGH. Trehalose/rhGH (1:4 mass ratio) and mannitol/rhGH (1:2 mass ratio) formulations (potassium phosphate buffer, pH 8.0) were lyophilized, reconstituted to 200 and 400 mg/mL rhGH, respectively, and then encapsulated in PLGA microspheres. The protein was released from these microspheres in its native state. Lyophilized formulations of rhGH yielded analogous results indicating the ability of trehalose and mannitol to stabilize the protein. Small solid particles of rhGH generated by spray drying (both air and freeze-drying) formulations containing Tween 20 or PEG were stable in ethyl acetate, but not methylene chloride. Similar results were also obtained with rhIFN-gamma (137 mg/mL in succinate buffer, pH 5.0), where both mannitol and trehalose were observed to stabilize the protein during exposure to the organic solvents resulting in the release of native rhIFN-gamma from PLGA microspheres. The rapid screening method allowed the development of stable concentrated protein solutions or solid protein formulations that could be successfully encapsulated in PLGA microspheres. The excipients observed to stabilize these proteins function by preferential hydration of the protein, and in the dry state (e.g., trehalose) may stabilize the protein via water substitution yielding a protective coating around the protein surface. Studies of other proteins should provide further insight into this mechanism of protein stabilization during encapsulation.