Understanding seasonal water budgets of wildlife is critical, as climate change alters water availability. Rocky Mountain mule deer (Odocoileus hemionus hemionus) mainly obtain water from their diet and drinking water; however, winter access to these resources varies with snow cover. To assess reliance on water sources by mule deer during winter, we longitudinally sampled female deer in Utah's Bear River Mountains, a montane region with deep snow and ample drinking water sources, and Pine Valley, a semiarid region with little snow. We implanted physiological biologgers to record heart rate and collected serum samples at the onset and end of winter over 2 yr. We measured the triple oxygen stable isotope composition of a body water sample (Delta ' O-17(BW)) distilled from serum. We then compared measured Delta ' O-17(BW) to predictions generated from two isotope models of water and oxygen fluxes that differed in their estimates of metabolic rate, based on either heart rate or body mass. Close agreement between measured and predicted Delta ' O-17(BW) validated our assumptions, although estimating metabolic rate via heart rate did not improve model predictions. Our results indicate that in the Bear River Mountains, deer increased their water drinking when snow cover limited foraging but that in Pine Valley, they relied on consistent food water input because of minimal snow cover. Estimates suggest that some deer in Pine Valley were seasonally independent from drinking water. Applying Delta ' O-17(BW) modeling enhances wildlife conservation by providing novel inferences about how individuals obtain water, informing proactive management in the face of climate change.