Purpose. The formulation and processing of protein drugs requires the stabilization of the native, biologically active structure. Our aim was to investigate the thermal stability of a model protein, lysozyme, in the presence of two model excipients, sucrose and hydroxypropyl-beta-cyclodextrin (HP-beta-CD). Methods. We used high sensitivity differential scanning calorimetry (HSDSC) in combination with a central composite design (CCD). As indicators of protein thermal stability, the measured responses were the unfolding transition temperature (T-m), the onset temperature of the denaturation (T-o), and the extrapolated onset temperature (T-o,T-c). Results. A highly significant (F probability <0.001) statistical model resulted From analysis of the data. The largest effect was due to pH lover the range 3.2-7.2), and the pH value that maximized T-m was 4.8. Several minor but significant effects were detected that were useful for mechanistic understanding. In particular, the effects of protein concentration and cyclodextrin concentration on T-m and T-o,T-c were found to be pH-dependent This was indicative of the partially hydrophilic nature of protein-protein interactions and protein-cyclodextrin interactions, respectively. Conclusions. Response surface methodology (RSM) proved efficient for the modeling and optimization of lysozyme thermal stability as well as for the physical understanding of the protein-sugar-cyclodextrin system in aqueous solution.