Habitat restoration efforts to conserve wildlife species are often conducted along a range of local site conditions, with limited information available to gauge relative outcomes for habitat suitability among sites and identify those that may lead to the greatest returns on restoration investment. We leveraged existing resource selection function models to generate heatmaps of spatially varying habitat suitability improvement potential for the Gunnison Sage-grouse (Centrocercus minimus) based on a suite of habitat restoration actions deployed across crucial habitats within six remaining satellite populations. We first simulated expected change in model covariates (habitat features) from a suite of restoration actions (increasing sagebrush, herbaceous, or litter cover, non-sagebrush shrub management, installation of mesic improvement structures, and removal of invasive plants) to generate modified input layers for each. We then reran the original models using these modified layers and calculated the predicted change in habitat suitability across space. The resulting heatmaps identify areas with the greatest improvement potential for each restoration action to help guide strategic restoration planning for the species. This data release, for the Poncha Pass satellite population, includes a set of 13 total raster files. These include: 6 uncategorized heatmaps illustrating predicted change in Gunnison Sage-grouse habitat suitability across space following habitat restoration actions (either single or combined), 6 categorized heatmaps additionally showing areas where 1) new habitat was created, 2) non-habitat remained non-habitat despite management interventions, or 3) negative changes in suitability were observed, and 1 heatmap illustrating predicted changes in suitability following new or worsening plant invasions (cheatgrass, represented by annual herbaceous). Habitat restorations vary by population depending on the reference model. We only ran management action simulations when the reference model had covariates suitable for the simulation (for example, pinyon juniper removal was only run when pinyon juniper was a covariate; See Saher and others (2022) for model details). Raster file names are coded as follows: PPb = Poncha Pass Breeding PPs = Poncha Pass Summer ahrb_rm = annual herbaceous removal (decrease in cover) ahrb_inv = annual herbaceous invasion (increase in cover) combo = combined actions mes_impr = mesic improvements (increase in area) pns_decr = decrease non-sagebrush shrub cover pns_incr = increase non-sagebrush shrub cover C = CATEGORIZED Maps V = UNCATEGORIZED Maps X = INVASION (categorized) Maps