Synthetic metabolism allows new metabolic capabilities to be introduced into strains for biotechnology applications. Such engineered metabolic pathways are unlikely to function optimally as initially designed and native metabolism may not efficiently support the introduced pathway without further intervention. To develop our understanding of optimal metabolic engineering strategies, a two-enzyme ethanol pathway consisting of pyruvate decarboxylase and acetaldehyde reductase was introduced into Synechocystis sp. PCC 6803. We characteriseda new set of ribosome binding site sequences in Synechocystis sp. PCC 6803 providing a range of translation strengths for different genes under test. The effect of ribosome-bindingsite sequence, operon design and modifications to native metabolism on pathway flux was analysed by HPLC. The accumulation of all introduced proteins was also quantified using selected reaction monitoring mass spectrometry. Pathway productivity was more strongly dependent on the accumulation of pyruvate decarboxylase than acetaldehyde reductase. In fact, abolishment of reductase over-expression resulted in the greatest ethanol productivity, most likely because strains harbouringsingle-gene constructs accumulated more pyruvate decarboxylase than strains carrying any of the multi-gene constructs. Overall, several lessons were learned. Firstly, the expression level of the first gene in anyoperon influenced the expression level of subsequent genes, demonstrating that translational coupling can also occur in cyanobacteria. Longer operons resulted in lower protein abundance for proximally-encoded cistrons. And, implementation of metabolic engineering strategies that have previously been shown to enhance the growth or yield of pyruvate dependent products, through co-expression with pyruvate kinase and/or fructose-1,6-bisphosphatase/sedoheptulose-1,7-bisphosphatase, indicated that other factors had greater control over growth and metabolic flux under the tested conditions.