The 32-ha Ruby Gulch Waste Rock Repository was created in 2001 to sequester heavy metal- and acid-generating rock exposed by past gold mining at Gilt Edge Mine in the northern Black Hills. Repository construction included the placement of a polyethylene membrane and a geotextile over waste rock to protect Ruby Gulch, a tributary of Strawberry Creek, from contaminated drainage. The membrane and geotextile were then covered with approximately 46 cm of drain rock, 76 cm of rocky subsoil and 15 cm of topsoil. The repository cap, comprised of ten erodible, 30-percent slopes and two plateaus, was seeded with a grass-forb mixture in 2003. The purpose of this study was to assess successional trends and long term sustainability of the vegetation established on the cap. Canopy cover, basal cover, and species diversity were measured annually in fifty-six 1-m² permanent plots and along twenty, 20-m permanent transects. Ocular estimations of cover by species were recorded for all 56, 1-m ² plots and 200, regularly spaced 0.25-m² microplots (10 per transect). An 8-point frame was used every 0.8 m along each transect to determine ground surface cover (200 points per transect/4000 points total). Data from 2005, 2006 and 2007 indicated a decrease in species richness, including a decline in broad-leaved plants (especially clovers, Trifolium spp.), and a near disappearance of thickspike wheatgrass (Elymus lanceolatus ). A few species remained stable, including western wheatgrass ( Pascopyrum smithii), Canada wildrye (Elymus canadensis), slender wheatgrass (Elymus trachycaulus), and Kentucky bluegrass (Poa pratensis). However, intermediate wheatgrass ( Thinopyrum intermedium), hard fescue (Festuca brevipila) and sheep fescue (Festuca ovina) significantly increased over the years. Severe drought and a grasshopper outbreak in 2006 appeared to accelerate compositional change between 2005 and 2006. Analyses of point frame data indicated an increase in vegetative ground cover as a result of significant increases in fescue, Kentucky bluegrass and litter. Nonmetric Multidimensional Scaling (NMS) ordination of transect cover data indicated increasing similarity in vegetative cover between sampling sites over the three years. Likewise, the cluster analysis based on each year’s permanent plot data demonstrated successional change with greater similarities among plots, resulting in fewer but larger groups over time. The results indicated long-term sustainability of vegetation with adequate cover to ensure erosion protection on the cap.