Abstract. Metabolic rate is a key aspect of organismal biology and the identification of selective factors that have led to species differences is a major goal of evolutionary physiology. We tested whether environmental characteristics and/or diet were significant predictors of interspecific variation in rodent metabolic rates. Mass-specific basal metabolic rates (BMR) and maximum metabolic rates (MMR, measured during cold exposure in a He-O 2 atmosphere) were compiled from the literature. Maximum (Tmax) and minimum (Tmin) annual mean temperatures, latitude, altitude, and precipitation were obtained from field stations close to the capture sites reported for each population (N ϭ 57). Diet and all continuous-valued traits showed statistically significant phylogenetic signal, with the exception of masscorrected MMR and altitude. Therefore, results of phylogenetic analyses are emphasized. Body mass was not correlated with absolute latitude, but was positively correlated with precipitation in analyses with phylogenetically independent contrasts. Conventional multiple regressions that included body mass indicated that Tmax (best), Tmin, latitude, and diet were significant additional predictors of BMR. However, phylogenetic analyses indicated that latitude was the only significant predictor of mass-adjusted BMR (positive partial regression coefficient, one-tailed P ϭ 0.0465). Conventional analyses indicated that Tmax, Tmin (best), and altitude explained significant amounts of the variation in mass-adjusted MMR. With body mass and Tmin in the model, no additional variables were significant predictors. Phylogenetic contrasts yielded similar results. Both conventional and phylogenetic analyses indicated a highly significant positive correlation between residual BMR and MMR (as has also been reported for birds), which is consistent with a key assumption of the aerobic capacity model for the evolution of vertebrate energetics (assuming that MMR and exercise-induced maximal oxygen consumption are positively functionally related). Our results support the hypothesis that variation in environmental factors leads to variation in the selective regime for metabolic rates of rodents. However, the causes of a positive association between BMR and latitude remain obscure. Moreover, an important area for future research will be experiments in all taxa are raised under common conditions to allow definitive tests of climatic adaptation in endotherm metabolic rates and to elucidate the extent of adaptive phenotypic plasticity.Key words. Bergmann's rule, body size, comparative method, diet, endothermy, energetics, temperature.Received August 29, 2003. Accepted February 8, 2004 Identification of the selective factors that have led to interspecific diversification in metabolic rates has been a major goal of ecological and evolutionary physiology (Garland and Carter 1994;Spicer and Gaston 1999;Walters and Reich 2000;McNab 2002). The most common approach is to correlate species or population differences in metabolic rate with variation in var...