The combination of directed mutagenesis with high-resolution structure analysis has made it possible to systematically address fundamental questions of protein folding and stability. Here we briefly review some recent results in this area based on studies of the lysozyme of bacteriophage T4. Extended segments of the polypeptide chain can be substituted with alanine, suggesting that about 50%, or perhaps less, of the overall amino acid sequence protein is necessary to define the 3-dimensional structure of the protein. It is the internal residues that seem to be most important for folding and stability (although not necessarily for function). Substitutions within the core of the protein of large nonpolar side chains with smaller ones have been used to better understand the nature of hydrophobic stabilization. Mutants that produce the largest cavities within the protein tend to be most destabilizing, allowing the energy cost of cavity formation to be estimated. Small, nonpolar ligands bind within such cavities and restore some stability to the protein. Analogous, nonpolar ligands do not bind, however, providing evidence that water molecules do not bind with high occupancy within nonpolar cavities. In a further series of studies it has been possible to re-engineer the active site region of T4 lysozyme to change the catalytic mechanism of the enzyme.