The analysis of protein residues recovered from archaeological artifacts provides a unique opportunity to reveal new information about past societies. However, many scientists are currently unwilling to accept protein-based results due to problems in method development and a basic lack of agreement regarding the ability of proteins to bind to, and preserve within, artifacts such as pottery. In this paper, I address these challenges by conducting a two-phase experiment. First, I quantitatively evaluate the tendency of proteins to sorb to ceramic matrices by using total organic carbon analysis and spectrophotometric assays to analyze samples of experimentally cooked ceramic. I then test a series of solvent and physical parameters in order to develop an optimized method for extracting and preparing protein residues for identification via mass spectrometry. Results demonstrate that protein strongly sorbs to ceramic and is not easily removed, despite repeated washing, unless an appropriate extraction strategy is used. This has implications for the future of paleodietary, conservation ecology and forensic research in that it suggests the potential for recovery of aged or even ancient proteins from ceramic matrices.