When a single host plant is infected by more than one strain of rhizobia, they face a tragedy of the commons. Although these rhizobia benefit collectively from nitrogen fixation, which increases host-plant photosynthesis, each strain might nonetheless increase its own reproduction, relative to competing strains, by diverting resources away from nitrogen fixation. Host sanctions can limit the evolutionary success of such rhizobial cheaters (strains that would otherwise benefit by fixing less nitrogen). Host sanctions have been shown in soybean (Glycine max) nodules, where the next generation of symbiotic rhizobia is descended from bacteroids (the differentiated cells that can fix nitrogen). Evidence for sanctions is less clear in legume species that induce rhizobial dimorphism inside their nodules. There, bacteroids are swollen and cannot reproduce regardless of how much nitrogen they fix, but sanctions could reduce reproduction of their undifferentiated clonemates within the same nodule. This rhizobial dimorphism can affect rhizobial evolution, including cheating options, in ways that may affect future generations of legumes. Both the importance of sanctions to hosts and possible physiological mechanisms for sanctions may depend on whether bacteroids are potentially reproductive.