Protein design is one of the great challenges of structural computational biology, with the potential to provide new catalysts, reagents and pharmaceuticals. Much progress has recently been made in design algorithms. However, structural characterization is essential for advancing computational protein design methods. Comparison of the predicted and experimentally determined structure validates design algorithms and reveals areas necessary for improvement. RosettaDesign, a computer program for protein design, was used to predict a low free energy sequence for a naturally occurring globular protein procarboxypeptidase (pdb code: 1AYE) given only its backbone coordinates. Chemical and thermal denaturation experiments have shown that the redesigned procarboxypeptidase (AYE) is more stable than the native protein. AYE resembles the native structure of procarboxypeptidase, but it forms a dimer at NMR concentrations (0.25-1.5μM). Based on chemical shift data from triple resonance NMR experiments, the Chemical Shift Index (CSI) program was used to validate the secondary structure of AYE. Further analysis of NMR data will be used to generate a high-resolution three-dimensional structure. Structural characterization of wild-type AYE will aid in future NMR experiments with mutated proteins. Further study of AYE would include eliminating dimerization and incorporating ligand-binding or catalytic sites. From studying AYE and other computer-generated proteins, we hope to gain more knowledge about the factors that determine protein structure, stability, and folding. This knowledge may ultimately be used to design novel proteins that have practical applications in industry and medicine.