The production of steroid and drug metabolites is a key area within the pharmaceutical industry. For a drug to pass safety profiling, large amounts of its metabolites are required to undergo extensive toxicology studies. Steroid compounds can be used in the treatment of chronic inflammatory disorders and also demonstrate efficacy in reducing cancerous tumour mass. A key step in the synthesis of these compounds is the insertion of an oxygen atom into a specific C-H unactivated bond. Cytochrome P450 (P450s) circumvent the limitations of synthetic methods for production of high value oxychemicals. The bacterial fatty acid monooxygenase P450 BM3 (CYP102A1) from Bacillus megaterium has the highest recorded P450 monooxygenase activity and is expressed in high yields in Escherichia coli. A plethora of P450 BM3 variants have demonstrated the proteinsâ amenability to mutations that alter its substrate binding profile towards non-natural substrates, leading to generation of valuable metabolites. This thesis provides novel information to add to the P450 BM3 knowledgebase. This work shows how a rational design approach, leading to the generation of a P450 BM3 A82F/F87V/I401P triple mutant (TM), results in the protein having a broad substrate selectivity profile for non-natural substrates. The TM BM3 enzyme catalyses the production of not only proton pump inhibitor (PPI) drug metabolites, such as 5-OH-omeprazole, but also oxidised steroid compounds such as 17-OH-progesterone. This is the first reported production of this key steroid pathway intermediate using bacterial P450 methods. This work demonstrates how only three influential point mutations can lead to activity with a wide range of novel substrates, through the destabilisation of the P450 BM3 protein. Further work details a different approach towards P450 BM3 engineering, where the aim was to produce two P450 BM3 variants that can generate 16Î±- and 16Î²-OH-testosterone. A multi-step screening effort identified two variants that produced both the desired products at >90% selectivity and conversion. The final piece of work in this thesis demonstrates how inclusion of a L86I mutation in the TM BM3 enzyme generates a quadrupule mutant (QM) (A82F/F87V/I401P/L86I) and leads to an overall increase in relative product formation for several tested substrates (including troglitazole, testosterone and omeprazole) and in some cases a change in product profile compared to the TM enzyme. This work led to the first reported production of OH-artemether, which has potential as an anti-malarial drug.