Unrestricted livestock grazing can degrade aquatic ecosystems, and its effects on native vertebrate species are generally mediated by changes to physical habitat. Recently, high estimated rates of cattle trampling on artificial redds within federal grazing allotments in southwestern Montana, USA, has raised concern that direct mortality from trampling may contribute to imperilment of native westslope cutthroat trout (Oncorhynchus clarkii lewisi). To explore the implications of cattle trampling, we built two mathematical models. First we used a temperature-driven model of egg-to-fry mortality representative of the developmental stages during which embryos would be vulnerable to trampling. Cattle trampling was an additional source of mortality (beyond natural mortality), and we modeled egg-to-fry mortality across a range of trampling rates (25–125% per month) for scenarios assuming low (0.60), moderate (0.81), and high (0.95) natural mortality. We then used a matrix model to determine how trampling affected population growth (λ), assuming initially stable (λ = 1.008) or slow-growing populations (λ = 1.025 and 1.05). Cattle trampling concentrated over a few days when the embryos were most sensitive caused greater egg-to-fry mortality than when the same amount of trampling occurred over one month. Trampling caused a large increase in egg-to-fry mortality when that natural mortality was low, but the overall population-level effect was far less than might have been anticipated from the rate of trampling itself. Nonetheless, small reductions in population growth rate could be biologically significant for populations with little or no demographic resilience, and trampling rates as low as 25% could lead to negative population growth. The rapid reduction in resilience with increased trampling rates (>50%) means that even growing populations are less likely to recover from periodic fluctuations. The overall risk posed by trampling will depend on whether cutthroat trout populations face concurrent threats that have already reduced their abundance and resilience. Biologists can potentially use the egg-to-fry model and thermograph data to identify dates when limiting cattle presence in or near stream habitats would likely reduce mortality from trampling. Evaluation of grazing policies on federal lands may be needed to ensure that species conservation and land use concerns are equitably balanced.