Species distribution modeling is increasingly used to describe and anticipate consequences of a warming ocean. These models often identify statistical associations between distribution and environmental conditions such as temperature and oxygen, but rarely consider the mechanisms by which these environmental variables affect metabolism. Oxygen and temperature jointly govern the rate of oxygen supply to oxygen demand, and theory predicts thresholds in these rates below which species population densities are diminished. However, parameterizing models with this joint dependence is challenging because of the paucity of experimental work for most species, and the limited applicability of experimental findings in situ. Here we ask whether the joint effects of temperature and oxygen can be reliably inferred from species distribution observations in the field, using the U.S. Pacific Coast as a model system. Through simulation testing, we found that our statistical model—which adapted the metabolic index to jointly consider oxygen and temperature by applying an Arrhenius equation and used a non-linear threshold function to link the index to fish distribution—could not precisely estimate the parameters due to inherent features of the distribution data. However, the model reliably estimated an overall metabolic index threshold effect, and provided a better fit to sablefish (Anoplopoma fimbria) spatial distribution than previously used models. This mechanistic approach may improve predictions of species distribution, even in novel environmental conditions. Further efforts to combine insights from mechanistic responses and realized species distributions will improve forecasts of species’ responses to future environmental changes.