Prairie restoration projects sometimes fail because of heavy invasion by invasive weeds, especially if they are not intensively managed. Few restoration projects are sampled after the first few years post-establishment, and little is known about what predictors are significant in maintaining restored communities over the very long term. Here, we stopped weeding experimental restoration plots to determine if persistence (that is, remaining unchanged after weeds are allowed to invade) of native prairie in western Iowa was related to planted species diversity