Rhizobia seem to have large degradative and metabolic capabilities that allow them to grow on diverse soil and rhizospheric substances, many of which are still unknown. Rhizobial genome sequences encode numerous transporters for unknown substrates, and transcriptomic studies have revealed genes with unknown functions that are highly expressed in roots or rhizospheres. It is proposed here that some of these unknown-function genes may have roles in the assimilation of root or soil substances and that rhizobial speciation avoids nutrient competition. Phenotypic tests, as currently performed in taxonomy (mainly for carbon and nitrogen usage), seem to underestimate rhizobial catabolic capabilities and the differences among species. Furthermore, considering that many Rhizobium transporter and catabolism genes are plasmid-borne, the value of phenotypic results in taxonomic studies is questionable. Genomotaxonomy could soon become a robust basis for proposing novel rhizobial species. © 2013 Elsevier GmbH.