Rhizobia - nitrogen-fixing, root-nodulating bacteria - play a critical role in both plant ecosystems and sustainable agriculture. Rhizobia form intracellular infections within legumes roots where they produce plant accessible nitrogen from atmospheric nitrogen and thus reduce the reliance on industrial inputs. The rhizobia-legume symbiosis is often treated as a pairwise relationship between single genotypes, both in research and in the production of rhizobial inoculants. However in nature individual plants are infected by a high diversity of rhizobia symbionts. How this diversity affects productivity within the symbiosis is unclear. Here, we use a powerful statistical approach to assess the impact of diversity within the Rhizobium leguminosarum - clover symbiosis using a biodiversity-ecosystem function framework. Statistically, we found no significant impact of rhizobium diversity. However this relationship was weakly positive - rather than negative - indicating that there is no significant cost to increasing inoculant diversity. Productivity was influenced by the identity of the strains within an inoculant; strains with the highest individual performance showed a significant positive contribution within mixed inoculants. Overall, inoculant effectiveness was best predicted by the individual performance of the best inoculant member, and only weakly predicted by the worst performing member. Collectively, our data suggest that the Rhizobium leguminosarum - clover symbiosis displays a weak diversity-function relationship, but that inoculant performance can be improved through the inclusion of high performing strains. Given the wide environmental dependence of rhizobial inoculant quality, multi-strain inoculants could be highly successful as they increase the likelihood of including a strain well adapted to local conditions across different environments.