1.Genetic variation for functionally important traits is ubiquitous in communities of nitrogen-fixing rhizobia, and while some studies have described significant effects of diversity on the functioning of plant-associated microbial communities, we lack a systematic test of how rhizobial diversity influences plant productivity. 2. The complexity of potential interactions among rhizobia and plants complicates the development of general predictions regarding causal relationships between rhizobial diversity and plant productivity. For example, while rhizobial complementarity may result in positive associations between symbiont diversity and plant productivity, antagonistic competition may reduce rhizobial community function. 3. Using two widespread native Australian Acacia species (A. salicina, A. stenophylla) and experimental rhizobial communities derived from 16 bacterial genotypes naturally associated with these hosts, we examined how the provision of mutualistic benefit varies with rhizobial identity, diversity and phylogenetic relatedness. 4. Analysis of plant performance in relation to rhizobial genotypic richness revealed that the presence of multiple rhizobial genotypes in the rhizosphere was associated with a general decrease in plant productivity compared to growth with single rhizobial genotypes. Importantly, these results appear to be robust in the face of variation in host identity and host diversity (i.e. one or two species mixtures). We also found that rhizobial genotypic identity and host species significantly influenced plant productivity in Acacia-rhizobia interactions, both in single and multi-strain inoculations. 5. Synthesis. Together, our data show that multiple rhizobia interacting with a single host species creates opportunities for emergent or higher-order effects that extend beyond those that could be simply predicted based upon outcomes of pairwise interactions, and that increased mutualist diversity does not necessarily translate into positive effects on plant growth.