BACKGROUND AND AIMS. Prairie restoration aims to create self-sustaining, resilient prairies that ameliorate biodiversity loss and soil deterioration associated with conversion of native grasslands to agriculture. Legumes are a key component of the nitrogen-limited prairie ecosystem. Evidence suggests that lack of suitable rhizobia may explain legume absence from restored prairies. This study explores effects of novel alternative inoculant delivery methods on:(a) prairie legume establishment, (b) soil biological properties, and (c) inoculant strain ability to nodulate the host over time. Methods:Alternative inoculation methods for seven legume species were tested in a replicated field experiment. Legume establishment, microbial biomass carbon and nitrogen, and rhizobial inoculant strain recovery were measured over a 3-year period. Results:Legume species richness in the second growing season was enhanced by a soil-applied granular clay inoculant, while seed-applied powdered peat inoculation was generally ineffective. When Dalea rhizobia were recovered 3-year after planting, only 2 % from the seed-applied inoculation treatment identified with the inoculant strains, whereas this amount ranged from 53 to 100 % in the other inoculation treatments. Some legumes established unexpectedly effective symbioses with strains not originally intended for them. Conclusions:Results provide new insights on inoculation of native legumes, especially when a mix of seeds is involved and the restoration targets harsh environments. © 2013 Springer Science+Business Media Dordrecht.