We provide underlying data for the publication "Plasmid design for tunable two-enzyme co-expression promotes whole-cell production of cellobiose". Please find the abstract below. Catalyst development for biochemical cascade reactions often follows a “whole cell approach” in which a single microbial cell is made to express all of the required enzyme activities. Although attractive in principle, the approach can encounter limitations when efficient overall flux from substrate to product necessitates precise balancing between the individual activities. Here, we show effective integration of major design strategies from synthetic biology to a coherent development of plasmid vectors enabling tunable two‐enzyme co‐expression in E. coli , for the purpose of whole‐cell production of cellobiose. Flux efficiency in the transformation of sucrose and glucose into cellobiose by a parallel (countercurrent) cascade of disaccharide phosphorylases requires the enzyme co‐expression cope with large differences in the specific activity of cellobiose phosphorylase (14 U mg−1) and sucrose phosphorylase (122 U mg−1). Comparing mono‐ and bicistronic co‐expression strategies, we analyze genetic elements controlling transcription, transcription‐translation coupling or plasmid replication for effect on activity, and also stable producibility, of the whole cell catalyst. We discover a key role of the bom (basis of mobility) site for plasmid stability dependent on the origin of replication and demonstrate the importance of RBS (ribosome binding site) strength for balanced bicistronic co‐expression. Whole cell catalysts show high specific rates (460 μmol cellobiose min−1 g−1 dry cells) and performance metrics (30 g L−1; ∼82% yield; 3.8 g L−1 h−1 overall productivity) promising for cellobiose production.