Evaluating the effectiveness of habitat management actions is critical to adaptive management strategies for conservation of imperiled species. We quantified the response of a Great Basin population of the Columbia spotted frog (Rana luteiventris) to multiple habitat improvement actions aimed to reduce threats and reverse population declines. We used mark-recapture data for 1,394 adult frogs that had been marked by state, federal, and university biologists in 9 ponds representing a single population over a 16-year period from 1997 to 2012. With the use of demographic models, we assessed population-level effects of 1) a grazing exclosure constructed around 6 stock ponds that had been used to water livestock for decades before being fully fenced in 2003, and 2) the construction of 3 new stock ponds in 2003 to provide alternative water sources for livestock and, secondarily, to provide additional frog habitat. These management actions were implemented in response to a decline of more than 80% in population size from 1997 to 2002. We found evidence that excluding cattle from ponds and surrounding riparian habitats resulted in higher levels of frog production (more egg masses), higher adult frog recruitment and survival, and higher population growth rate. We also found that frogs colonized the newly constructed stock ponds within 3 years and frogs began breeding in 2 of them after 5 years. The positive effects of the cattle exclosure and additional production from the new ponds, although notable, did not result in full recovery of the population even 9 years later. This slow recovery may be partly explained by the effects of weather on recruitment rates, particularly the negative effects of harsher winters with late springs and higher fall temperatures. Although our findings point to potential successes of habitat management aimed at slowing or reversing rapidly declining frog populations, our study also suggests that recovering from severe population declines can take many years because of demographic and environmental processes.