In the legume-rhizobium mutualism, leguminous plants provide rhizobia with sugars and a root nodule environment conducive to nitrogen fixation. In return, plants receive reduced nitrogen from these symbiotic bacteria. The symbiosis is established anew each plant generation when rhizobia residing in the soil infect the roots of germinating plants and differentiate into nitrogen-fixing bacteroids within the nodule cells of the host plant, before passing back into the soil at plant senescence. Therefore, plants and rhizobia experience both symbiotic and a free-living lifestyles and may exhibit adaptations specialized to the particular challenges of either lifestyle. This dissertation examines how environmental context across contrasting soil types affects adaptive differentiation in legumes and rhizobia both in symbiosis and in free-living life stages. The first chapter examines whether legumes and rhizobia in symbiosis exhibit adaptive differentiation across soil conditions. Co-invading exotic legumes and rhizobia have expanded their range to include harsh serpentine and more benign non-serpentine soil in a California grassland. A greenhouse experiment demonstrates that the ability to allocate to symbiotic tissue in a specific soil context may be a key adaptation underlying legume differentiation across heterogeneous environments during a biological invasion. The second chapter examines whether free-living rhizobia exhibit adaptive differentiation across soil types. The assemblages of rhizobia associated with native and invasive host legumes in both harsh, nickel-laden serpentine and more benign nonserpentine soil were evaluated for soil-specific differentiation to the presence or absence of nickel. Native rhizobia display adaptive differentiation, with serpentine strains outperforming non-serpentine strains in the presence of nickel and non-serpentine strains outperforming serpentine strains in the absence of nickel. However, invasive rhizobia do not display adaptive divergence, despite exhibiting an evolutionary trade-off between nickel tolerance and growth rate similar to that found in the native strains. The third chapter examines whether free-living legume seeds exhibit adaptive differentiation across soil environments. The seeds of native legume populations display striking variability in coloration, as do the soils these populations inhabit. A quantitative analysis of heritable seed and soil colors indicates a strong concordance between seed and home soil color. Seed predation appears to have driven selection for locally camouflaged seeds, maintaining adaptive differentiation in seed pigmentation.