Use of in silico modeling technologies to guide rational strain engineering can greatly accelerate the pace of industrial bioprocess development. One such technique is the OptKnock approach, which identifies a set of gene deletions designed to couple product formation to growth; i.e., in order for the cell to grow efficiently, it must also produce the compound of interest. A complementary experimental approach, evolutionary engineering, uses controlled selection pressure to optimize strain fitness and growthrate following genetic manipulations. In addition to achieving superior product yield, strains generated by this approach are suitable for continuous bioprocessing, due to their inherent genetic stability. In this work, OptKnock was used to design two E. coli strains for the growth-coupled production of succinic acid. The first strain, AB3, contained 3 gene deletions and had anaerobic succinate yield on glucose of 21 mole %, approximately 5 times that of wild-type E. coli. The second, AB4, contained 5 gene deletions and produced a yield of 51 mole %. For further improvement, these strains were subjected to adaptive evolution using a specially designed machine to provide frequent serial dilutions. For example, 30 days of evolution increased product yield of AB3 to 113 mole %, which is very close to the maximum predicted for the strain. The engineered strains were finally tested in continuous fermentation to demonstrate superior volumetric productivity as well as genetic stability. The combined computational and experimental approach presented here has the potential to significantly expedite the metabolic engineering of industrial organisms for continuous bioprocess applications.