Despite a U.S. goal of ‘no-net-loss’ of wetland structure and function, restoration performance standards are typically based on limited criteria, with soil carbon, nutrient, and microbial criteria being particularly rare. We examined plant community composition, diversity, and various soil functional variables for two different restoration techniques, topsoil-removal and solarization, in wetland prairies in Oregon, USA. We compared three site-level replicates of each treatment to three high-quality remnant wetland prairies. We found significant tradeoffs among diversity, native cover, and ecosystem functioning between restoration treatments. Wetlands with topsoil removed had similar diversity to remnants, but lower productivity, soil carbon and nitrogen, microbial biomass, and arbuscular mycorrhizal fungal infection rates. Solarization sites had the highest native cover, but diversity was approximately half that of remnant wetlands. We attribute this to both a priority effect of seeded native perennial bunchgrasses establishing early in the restoration process and competitively excluding other species, and to a lack of microtopographic variation in the restored sites. Our results suggest that restoration projects should evaluate both structural and functional processes, since they may reveal tradeoffs among important goals. Mitigation efforts should strive to understand the mechanisms causing these tradeoffs among structure and function and try to minimize these in restoration designs.