Basal metabolic rate (BMR) is probably the most studied aspect of energy metabolism in vertebrate endotherms. Numerous papers have explored its mass allometry, phylogenetic and ecological relationships, and ontogeny. Implicit in many of these studies (and explicit in some) is the view that BMR responds to selection, which requires repeatability and heritability. However, BMR is highly plastic in response to numerous behavioral and environmental factors and there are surprisingly few data on its repeatability. Moreover, the mechanistic underpinnings of variation in BMR are unclear, despite considerable research. We studied BMR repeatability in deer mice (Peromyscus maniculatus) across intervals of 30-60 days, and also examined the influence of birth altitude (3,800 m versus 340 m) and temperature acclimation (to approximately 5 or approximately 20 degrees C) on BMR, and the relationship between BMR and organ size. Neither acclimation temperature nor natal altitude alone influenced BMR, but the combination of birth at high altitude and cold acclimation significantly increased BMR. Few visceral organ masses were correlated to BMR and most were inconsistent across natal altitudes and acclimation temperatures, indicating that no single organ 'controls' variation in BMR. In several treatment groups, the mass of the 'running motor' (combined musculoskeletal mass) was negatively correlated to BMR and the summed mass of visceral organs was positively correlated to BMR. We found no repeatability of BMR in any treatment group. That finding-in sharp contrast to high repeatability of BMR in several other small endotherms-suggests little potential for direct selection to drive BMR evolution in deer mice.