Prairie reconstructions are a critical component of preservation of the imperiled tallgrass prairie ecosystem in the Midwestern United States. Sustainability of this endeavor depends on establishment of persistent cover of planted native species and resistance to noxious weeds. The goal of this study was to understand the influence of early reconstruction practices on long-term outcomes. Twelve replicates of three planting methods (dormant-season broadcast, growing-season broadcast, and growing-season drill) and three seed mix richness levels (10, 20, or 34 species), fully crossed in a completely randomized design were planted in 2005 on nine former agricultural fields located in Iowa and Minnesota. Cover by species was estimated in 2005–2007, 2010, and 2015. In 2015, cover of planted species, native nonplanted species, and exotic species were similar to those recorded in 2010. Cover of the noxious weed Cirsium arvense had also declined by an average of 49% without herbicide from a peak in 2007 to low stable levels from 2010 to 2015. Richness of planted forbs, on the other hand, were still increasing in high-richness broadcast treatments (e.g. 17–59% increase 2010–1015 in Minnesota). Two results in 2015 are reasons for concern: cover of planted species is only slightly over 50% in both Minnesota and Iowa, though with forbs still increasing, this may improve; and the cool-season exotic grasses Poa pratensis and Bromus inermis are increasing at both Minnesota and Iowa sites. Control of these invasive grasses will be necessary, but care will be needed to avoid negative impacts of control methods on natives.