Abstract: Isoprenoids are found in such diverse commercial products as pharmaceuticals, perfumes, and foods. In this thesis work, I addressed two limitations to the ability to produce isoprenoids in Escherichia coli. First, the poor expression of plant terpene synthases in E. coli restricted production to a few nanograms per milliliter of culture. We were unable to produce enough active enzyme to consume the substrate supplied by E. coli's biosynthetic pathway. We were also limited in our coverage of possible isoprenoid products. Prior to this work, terpene synthase discovery relied on conserved sequence motifs among plant terpene synthases, limiting the scope of terpene synthase discovery. We hypothesized that the poor performance of plant terpene synthase genes in E. coli was related to the frequency of rare codons in the genes. To overcome this limitation, I synthesized two codon-optimized genes coding for amorphadiene synthase (ADS) and epi-cedrol synthase (EPC). The expression of ADS resulted in the first reported production of a precursor to the anti-malarial artemisinin in a heterologous host. Comparison of the native and synthetic EPC showed a 9-fold improvement in epi-cedrol production. A nearly 2-fold further improvement in epi-cedrol production was realized through directed evolution of the two synthetic genes. I also developed a new screen for terpene synthase activity. This screen is based on an observation that elevated levels of the precursors to isoprenoids inhibit the growth of E. coli. This screen allowed us to enrich a population of clones for those capable of consuming the precursors. Because the screen relies on catalytic activity rather than sequence, it should apply to any DNA library encoding functional enzymes. A variety of libraries was screened in order to demonstrate broad applicability. I screened a library of genomic DNA fragments from Bacillus subtilis strain 6051, a known isoprene producer. This resulted in two genes coding for proteins with activity towards the isoprene precursor DMAPP. I have also enriched a library of mutant terpene synthases, resulting in the detection of promising terpene synthase activities within the enriched population. I have also screened cDNA libraries from Artemisia annua and Saccharomyces cerevisiae.