Microbial production of mevalonate from renewable feedstock is a promising and sustainable approach for the production of value-added chemicals. We describe the metabolic engineering of Escherichia coli to enhance mevalonate production from glucose and cellobiose. First, the mevalonate-producing pathway was introduced into E. coli and the expression of the gene atoB, which encodes the gene for acetoacetyl-CoA synthetase, was increased. Then, the deletion of the pgi gene, which encodes phosphoglucose isomerase, increased the NADPH/NADP(+) ratio in the cells but did not improve mevalonate production. Alternatively, to reduce flux toward the tricarboxylic acid cycle, gltA, which encodes citrate synthetase, was disrupted. The resultant strain, MG Delta gltA-MV, increased levels of intracellular acetyl-CoA up to sevenfold higher than the wild-type strain. This strain produced 8.0 g/L of mevalonate from 20 g/L of glucose. We also engineered the sugar supply by displaying beta-glucosidase (BGL) on the cell surface. When cellobiose was used as carbon source, the strain lacking gnd displaying BGL efficiently consumed cellobiose and produced mevalonate at 5.7 g/L. The yield of mevalonate was 0.25 g/g glucose (1 g of cellobiose corresponds to 1.1 g of glucose). These results demonstrate the feasibility of producing mevalonate from cellobiose or cellooligosaccharides using an engineered E. coli strain.