Range of different subtilisins are recombined within the context of a common core backbone donated by Savinase to generate new protein variants. Up to 6 functionally rich regions of Savinase were replaced with the corresponding regions from 6 different bacilli subtilisins. Savinase was largely tolerant to a wide variety of different region combinations Z-VAD-FMK despite a sequence identity of #60% with the other subtilisins. The substrate specificity of the hybrid variants varied and depended on the sequence combinations at each region. To allow effective shuffling of important functional regions of subtilisins, a combinatorial fragment exchange approach was employed. Such an approach will allow sequence homology-independent recombination of multiple regions of a variety of different sequences originating from homologues with varying degrees of sequence identity to the core target protein. Loop regions for example can play pivotal roles in defining functional differences between protein homologues yet their sequence, length and structure can vary dramatically. Furthermore, loop regions are more adaptable than regular secondary structure and thus accommodate replacement more readily. Thus a loop region, which can be precisely defined by analysis of a protein structure or model, may represent a recombination unit. Precise structural alignments allow the exact sequence region to be exchanged thus overcoming sequence anomalies due to insertions or deletions or inherent low sequence homology. Inclusion of insertion and deletions will allow a broader sampling of conformational space than that accessed by substitutions alone; indel mutation have been found to be beneficial in altering properties such as substrate specificity. Seven different subtilisins were chosen to act as sequence donors. Sav was chosen as the core protein due to its general robustness, as highlighted by its widespread usage in different commercial applications far removed from its normal biological context. Three other commonly studied and utilised bacterial subtilisins were also chosen to contribute diversity; AlcalaseH, BPN’, and Subtilisin E. Two subtilisins derived from thermophiles were also included; Thermitase and Bacillus Ak.1. They also contain a novel calcium binding site that is thought to contribute stability. Furthermore, AK1 has a relatively uncommon CysX-Cys disulphide bridge separated by only one residue that contributes to thermostability and formation of the substrate binding. While the detailed 3D structure of the above selected subtilisins were known, there was no structure available before the commencement of this work for the final sequence contributor, the intracellular subtilisin protease. The ISPs are the only members of the bacterial subtilisins that function within the cell and have sequence features that differ from their secreted relatives, including the absence of the classical prodomain. The sequence identity of ISP to Sav was also the lowest amongst the selected homologues. Therefore, for the purposes of this work a structural model of an ISP from B. clausii was generated. Comparison of the homology model with the recently determined 3D structure of the ISP revealed.