The antifungal polyketide soraphen A is produced by the myxobacterium Sorangium cellulosum So ce26. The slow growth, swarming motility and general intransigence of the strain for genetic manipulations make industrial strain development, large-scale fermentation and combinatorial biosynthetic manipulation of the soraphen producer very challenging. To provide a better host for soraphen A production and molecular engineering, the biosynthetic gene cluster for this secondary metabolite was integrated into the chromosome of Streptomyces lividans ZX7. The upstream border of the gene cluster in Sor. cellulosum was defined by disrupting sorC, which is proposed to take part in the biosynthesis of methoxymalonyl-coenzyme A, to yield a Sor. cellulosum strain with abolished soraphen A production. Insertional inactivation of orf2 further upstream of sorC had no effect on soraphen A production. The genes sorR, C, D, F and E thus implicated in soraphen biosynthesis were then introduced into an engineered Str. lividans strain that carried the polyketide synthase genes sorA and sorB, and the methyltransferase gene sorM integrated into its chromosome. A benzoate-coenzyme A ligase from Rhodopseudomonas palustris was also included in some constructs. Fermentations with the engineered Str. lividans strains in the presence of benzoate and/or cinnamate yielded soraphen A. Further feeding experiments were used to delineate the biosynthesis of the benzoyl-coenzyme A starter unit of soraphen A in the heterologous host.