Oil sands mining disturbs thousands of hectares of boreal landscape, about 65% of which is wetland. Its reclamation will constitute the largest wetland reclamation project in Canadian history. We developed a unified analytical framework that we used to set reclamation targets and evaluate reclamation progress using submersed aquatic vegetation (SAV). We sampled SAV in 38 minimally disturbed wetlands to establish a reference condition and compared this to SAV in 25 reclamation wetlands. We observed 26 taxa: all were native and five are regionally rare. Using a combination of ordination, clustering, and indicator species analysis, we identified seven SAV assemblages, distinguishable based on 10 indicator species. The assemblages found in wetlands contaminated by tailings had significantly fewer taxa. Using joint plots, we demonstrate that they differ in terms of environmental variables reflecting depth, slope, salinity, transparency, water and sediment nutrient levels, and alkalinity. Collectively, 74% of reference wetlands had SAV belonging to either marsh or fen–marsh assemblages that we consider suitable targets for reclamation. Having multiple reclamation targets avoids creating a homogenous post-mining landscape with low gamma diversity. Using complementary multivariate and categorical tests, we found that reclamation wetlands failed to support either of these dominant reference assemblages. Instead, every reclamation wetland supported one of five atypical SAV assemblages, each of which was uncommon (<10%) among reference wetlands. The analytical framework we developed can be applied to other ecosystems or taxa to develop targets for reclamation, to evaluate reclamation progress, and to identify environmental characteristics associated with reclamation targets.