1.Understanding the mechanisms driving geographic range sizes of species is a central issue in ecology, but remarkably few rules link physiology with the distributions of species. Maximal metabolic rate (MMR) during exercise is an important measure of physiological performance. It sets an upper limit to sustained activity and locomotor capacity, so MMR may influence ability to migrate, disperse, and maintain population connectivity. Using both conventional ordinary least squares (OLS) analyses and phylogenetically generalized least squares (PGLS), we tested whether MMR helps explain geographic range size in 51 species of placental mammals. 2.Log body mass alone (OLS r2 = 0.074, p = 0.053; PGLS r2 = 0.016, p = 0.373) and log MMR alone (OLS r2 = 0.140, p = 0.007; PGLS r2 = 0.061, p = 0.081) were weak predictors of log range size. 3.However, multiple regression of log body mass and log MMR accounted for over half of the variation in log range size (OLS R2 = 0.527, p < 0.001). The relationship was also strong after correcting for the phylogenetic non-independence (PGLS R2 = 0.417, p < 0.001). 4.In analyses restricted to rodents (34 species), neither log body mass alone (OLS r2 = 0.004, p = 0.720; PGLS r2 = 0.003, p = 0.77) nor log MMR alone was useful in predicting log geographic range size (OLS r2 = 0.008, p = 0.626; PGLS r2 = 0.046, p = 0.225), but multiple regressions of log body mass and log MMR accounted for roughly a third to a half of the variation in log range size (OLS R2= 0.443, p < 0.001, PGLS r2 = 0.381, p < 0.001). 5.Mass-independent MMR is a strong predictor of mass-independent geographic range size in placental mammals. The ability of body mass and MMR to explain nearly 50% of the variation in the geographic ranges of mammals is surprising and powerful, particularly when neither variable alone is strongly predictive. 6.A better understanding of MMR during exercise may be important to understanding the limits of geographic ranges of mammals, and perhaps other animal groups.