In recent decades, feral horse (Equus caballus; horse) populations increased in sagebrush (Artimesia spp.) ecosystems, especially within the Great Basin, to the point of exceeding maximum appropriate management levels (AML_max), which were set by land administrators to balance resource use by feral horses, livestock, and wildlife. Concomitantly, greater sage-grouse (Centrocercus urophasianus; sage-grouse) are sagebrush obligates that have experienced population declines within these same arid environments as a result of steady and continued loss of seasonal habitats. Although a strong body of research indicates that overabundant populations of horses degrade sagebrush ecosystems, empirical evidence linking horse abundance to sage-grouse population dynamics is missing. Within a Bayesian framework, we employed state-space models to estimate population rate of change (λ) using 15 years (2005–2019) of count surveys of male sage-grouse at traditional breeding grounds (i.e., leks) as a function of horse abundance relative to AML_max and other environmental covariates (e.g., wildfire, precipitation, % sagebrush cover). Additionally, we employed a post hoc impact-control design to validate existing AML_max values as related to sage-grouse population responses, and to help control for environmental stochasticity and broad-scale oscillations in sage-grouse abundance. On average, for every 50% increase in horse abundance over AML_max, our model predicted an annual decline in sage-grouse abundance by 2.6%. Horse abundance at or below AML_max coincided with sage-grouse λ estimates that were consistent with trends at non-horse areas elsewhere in the study region. Thus, AML_max, as a whole, appeared to be set adequately in preventing adverse effects to sage-grouse populations. Results indicated 76%, 97%, and >99% probability of sage-grouse population decline relative to controls when horse numbers are 2, 2.5, and ≥3 times over AML_max, respectively. As of 2019, horse herds exceeded AML_max in Nevada, USA, by >4 times on average across all horse management areas. If feral horse populations continue to grow at current rates unabated, model projections indicate sage-grouse populations will be reduced within horse-occupied areas by >70.0% by 2034 (15-year projection), on average compared to 21.2% estimated for control sites. A monitoring framework that improves on estimating horse abundance and identifying responses of sage-grouse and other key indicator species (plant and animal) would be beneficial to guide management decisions that promote co-occurrence of horses with sensitive wildlife and livestock within landscapes subjected to multiple uses. Published 2021. This article is a U.S. Government work and is in the public domain in the USA. The Journal of Wildlife Management published by Wiley Periodicals LLC on behalf of The Wildlife Society.