Polyketides are a class of small molecules synthesized by a broad spectrum of bacteria, plants, and fungi, and many exhibit powerful bioactive properties. The number of clinically-relevant compounds adapted from polyketide scaffolds is growing, eliciting attempts from synthetic organic chemists to construct polyketide-related compounds in the laboratory from simple chemical building blocks. Unfortunately, the current efficiency by which a skilled artisan can synthesize even small quantities of a polyketide is severely limited by the functional and stereochemical complexity of these compounds. Conceptually, it would be much simpler to genetically reprogram the enzymes responsible for polyketide biosynthesis to produce designer molecules; however, the massive size of polyketide synthase enzymes has hindered efforts towards understanding critical features of their structures and mechanisms. Only very recently has structural information become available for enzymes involved in polyketide biosynthesis, providing an initial glimpse into the inner workings of these subcellular pharmaceutical factories. It will not be possible for mankind to fully realize the potential of engineered polyketide synthases without understanding how their architectures govern the molecules they have evolved to produce. In this work, the structure and mechanism of several enzymes involved in polyketide biosynthesis is investigated. An unprecedented architecture for the ketoreductase-enoylreductase didomain from the second module of the spinosyn polyketide synthase reveals structural divergence from the related mammalian fatty acid synthase, and reconstituted in vitro activity of the enoylreductase domain indicates the isolated enzyme retains activity apart from its parent polyketide synthase module. The dehydratase domain isolated from the tenth module of the rifamycin polyketide synthase, previously hypothesized to only form double bonds with (Z) geometry, was found to have altered stereoselectivity dependent on the carrier handle bound to the substrate. The enoyl-isomerase domain, isolated from the fourteenth module of the bacillaene polyketide synthase, utilizes a catalytic mechanism that relies only on a single active site histidine. A series of ketosynthase domains from trans-acyltransferase polyketide synthases reveal how polyketides bind covalently to the active site of the ketosynthase, and how the flanking subdomain of the ketosynthase is used as an anchor point for the assembly of the polyketide synthase megacomplex.