This study describes the realization of an anoxic acetoin production process using the proteobacterium Shewanella oneidensis. Fermentative processes are of high biotechnological relevance since they offer high productivity and a low percentage of substrate consumption for anabolic processes. Nevertheless, the range of compounds that can be produced as sole end product of a fermentative process is limited, since the average oxidation state of substrate and products has to be identical in the absence of an external electron acceptor. This limitation could be overcome by the transfer of the surplus of electrons to a poised electrode surface, which of note is the only known anaerobic electron acceptor that cannot be depleted. In the first genetic engineering step, deletion mutants were developed that are devoid of either one, two, or all three prophages in their genome with the aim to construct a more stable chassis strain for microbe-electrode interaction, due to less prophage induced cell lysis (Gödeke et al., 2011). Current production in a bioelectrochemical system together with the analysis of cells on the anode surface were used as surrogate for the stability assessment. The λ-prophage deletion mutant produced overall 1.34fold more current (6.7 μA cm ) than the wild type and all other constructed strains and showed with 1.1 × 10 cells the highest cell density on the anode surface (2.3fold more than the wild type). The strain was further modified to contain codon optimized versions of acetolactate synthase and acetolactate decarboxylase derived from Bacillus subtilis. This allowed for the production of a mixture of acetoin and acetate from lactate in an almost 0.4:1 ratio. Further process improvement was reached by the deletion of the acetate kinase and phosphotransacetylase genes ackA/pta. The acetoin yield increased in this mutant from 40 to 86% of the theoretical maximum and acetoin was the only detectable end product. Biotechnol. Bioeng. 2017;114: 1283-1289. © 2017 Wiley Periodicals, Inc.