Grassland to shrubland transitions are well documented throughout the desert grassland region of the Chihuahuan Desert. These transitions were triggered in the early 20th century by overgrazing of perennial grasses during drought periods, loss of fire regimes, and seed dispersal by livestock. Shrublands dominated by C3 species are highly resilient to natural disturbances, so herbicides specific to the C3 photosynthetic pathway are used to reduce the competitive preemption of soil water resources by shrubs and trigger the recovery of C4 perennial grasses. Anecdotally, there is evidence that shrub management treatments have been effective in some cases, but not in others, creating controversy about their value. As part of the Restore New Mexico partnership, we initiated a region-scale, long-term monitoring experiment embedded within shrub management treatments beginning in 2007. Within shrub management areas, we created treatment and control plots that were matched according to soils/ecological sites and initial state prior to treatments, such that control plots were excluded from treatment. Line-point intercept was used to measure cover prior to and after treatments. In addition, we sampled older treatments (2004 and earlier) using space-for-time substitution assumptions. Our records to date indicate that 1) treatments are generally effective in increasing perennial grass cover, 2) the composition of grasses that increase usually differs from that of putative reference states, creating “novel ecosystems” and 3) soil and climate properties are related to the magnitude of grass response to treatments; specifically, loamy soils and drier climates yield reduced grass response. Experimental monitoring data can be used to refine restoration approaches and resolve controversies about them.