We recently discovered that PhlD, a type III polyketide synthase in the diacetylphloroglucinol (DAPG) biosynthetic pathway, is in fact a phloroglucinol synthase. This finding creates the foundation for commercial biosynthesis of phloroglucinol either in vivo from glucose via fermentation by E. coli, or in vitro by combining substrate with purified protein or cell lysate. Characterization of the biochemical properties of PhlD revealed that like many enzymes from secondary metabolism, it suffers from slow catalytic activity and poor stability. Therefore we are performing directed evolution to engineer PhlD to become a practical industrial biocatalyst. Toward this end, a colorimetric screening system to quantify phloroglucinol production was developed and used to evaluate a library of PhlD mutants. In a first round of screening, the PhlD mutant library was generated by synthetic shuffling of the diversity present in 50 natural PhlD homologs. After screening approximately 40,000 clones, two significantly improved PhlD variants were discovered which possessed increases in both catalytic efficiency and thermostability. Based on these promising results, a second round of directed evolution is underway. We will apply the THR selection method, recently developed by Delcourt and coworkers, to further increase the thermostability of PhlD in an activity-independent approach by fusing PhlD to a thermostable kanamycin resistance gene and expressing the fusion protein in Thermus thermophilus.