The work in this thesis concerns the biosynthetic pathways to antibiotic natural products in actinomycete bacteria, and evaluates three different approaches to accessing novel chemical diversity from these pathways. In Chapter 1, major pathways of natural product biosynthesis in Streptomyces and allied bacteria are discussed, and methods for manipulating them to produce new derivatives. Chapter 2 sets out the Materials and Methods used in this work. Chapter 3 concerns the pathway to the giant linear polyol-polyene clethramycin from Streptomyces malaysiensis DSM4137, a rare sulfonated natural product. Gene deletion was used to identify the essential sulfotransferase involved. In Chapter 4, a novel genetic engineering approach dubbed "accelerated evolution" was explored with the clethramycin PKS, which had previously been used to generate multiple analogues of the polyketide rapamycin. In this approach, recombination within the homologous modules of a PKS is stimulated and the resulting progeny are examined for the production of novel compounds. Unfortunately, the approach did not prove useful here, and possible reasons for this are discussed. In Chapter 5, genome mining was used to explore the rare genus Saccharopolyspora which has previously provided the valuable erythromycin A and spinosyn. The genomes of 20 strains of Saccharopolyspora were analysed, revealing over 400 gene clusters, including novel examples of valuable chemotypes. In Chapter 6, the gene cluster was characterised from Saccharopolyspora flava for one of these novel compounds, a macrocyclic polyene named here flavatericin. Conditions were found under which Sa. flava produced multiple flavatericins. The starter unit was identified as 3 hydroxybenzoic acid by expression and assay of the chorismatase gene ftcR. Deletion of ftcR abolished flavatericin production. One flavatericin component was purified to homogeneity and subject to NMR analysis. Chapter 7 summarises conclusions, and outlines suggestions for future work.