Escherichia coli and Saccharomyces cerevisiae are two model organisms whose genes and proteins have been well studied experimentally, including the complete sequencing of their genomes.  Their small molecule metabolic pathways are thoroughly characterised, so the sets of enzymes that make up these pathways are known.  We were able to determine the domain architecture of these proteins through sequence homology to proteins of known structure as well as by assigning sequence domains of unknown structure.  Thus 85% of the yeast enzymes and 88% of the E. coli enzymes were assigned a structural or sequence domain. The domain assignments allowed us to answer general questions about the construction of small molecule metabolic pathways within a single organism, E. coli.  Just under half of the roughly 600 E. coli small molecule metabolic enzymes are multi-domain proteins.  Their domains belong to just over 200 protein families, so that duplication has been an important evolutionary mechanism in this protein network.  Tracing the duplications and combinations of domains in the different enzymes in this network provides insights into the evolution of small molecule metabolism.  We were also able to compare the E. coli enzymes and pathways to yeast small molecule metabolic pathways in order to investigate the extent of conservation in enzymes and pathways after roughly three billion years of evolution.