When calorically restricted at cool ambient temperatures, mice conserve energy by entering torpor, during which metabolic rate (MR), body temperature (T(b)), heart rate (HR), and locomotor activity (LMA) decrease. Treatment with exogenous adenosine produces a similar hypometabolic state. In this study, we conducted a series of experiments using the nonspecific adenosine receptor antagonists aminophylline and 8-sulfophenyltheophylline (8-SPT) to test the hypothesis that adenosine signaling is necessary for torpor in fasted mice. In the first experiment, mice were subcutaneously infused with aminophylline while T(b), HR, and LMA were continuously monitored using implanted radiotelemeters. During a 23-h fast, saline-treated mice were torpid for 518 ± 43 min, whereas aminophylline-treated mice were torpid for significantly less time (54 ± 20 min). In a second experiment, aminophylline was infused subcutaneously into torpid mice to test the role of adenosine in the maintenance of torpor. Aminophylline reversed the hypometabolism, hypothermia, bradycardia, and hypoactivity of torpor, whereas saline did not. In the third and fourth experiments, the polar adenosine antagonist 8-SPT, which does not cross the blood-brain barrier, was infused either subcutaneously or intracerebroventricularly to test the hypothesis that both peripheral and central adenosine receptor signaling are necessary for the maintenance of torpor. Intracerebroventricular, but not subcutaneous, infusion of 8-SPT causes a return to euthermia. These findings support the hypothesis that adenosine is necessary for torpor in mice and further suggest that whereas peripheral adenosine signaling is not necessary for the maintenance of torpor, antagonism of central adenosine is sufficient to disrupt torpor.