Blood pressure is usually well maintained during epidural or spinal anesthesia even in the presence of extensive sympathetic blockade. The authors investigated whether hormonal systems support arterial pressure and how the circulation copes with a hypoxic challenge when activation of the sympathetic nervous system is selectively impaired by neural blockade. Accordingly, the effects of high epidural anesthesia alone and combined with hypoxia were evaluated in seven awake trained dogs. On different days, either bupivacaine 0.5% (8–12ml) or saline (placebo) were randomly injected epidurally and the effects evaluated on cardiovascular (arterial pressure, heart rate) and respiratory (blood gases, oxygen consumption) variables, as well as on hormone plasma concentrations (vasopressin, norepinephrine, epinephrine, renin) during both normoxia and hypoxia. During epidural anesthesia alone, vasopressin increased tenfold (1.7 pg/ml ± 1.0 SD to 16.8 ± 13.8,P< 0.05), norepinephrine decreased (90 pg/ml ± 31 to 61 ± 28,P< 0.05) while epinephrine and renin concentrations remained unchanged. Mean arterial and pulse pressure decreased by 13 mmHg and 23 mmHg (P< 0.05), respectively. In dogs without sympathetic blockade (saline group), hypoxemia (Pa&OV0312;: 31 ± 4 mmHg) evoked an increase in mean blood pressure by 37 mmHg ± 8 and heart rate by 50 beats per min ± 17. In contrast, in the presence of sympathetic blockade but with a similar degree of hypoxemia, blood pressure failed to increase (+1 mmHg ± 14) and heart rate rose by only 15 beats per min ± 11. These differences between groups were statistically significant (P< 0.001). Hypoxemia induced a similar hypocarbia (Pa&OV0312;: 25 mmHg) in both groups, indicating that the ventilatory response to hypoxemia was preserved after epidural blockade. During hypoxemia vasopressin concentrations increased 35-fold to 64 pg/ml ± 38 (P< 0.0001) compared to base line only during epidural anesthesia, but not after epidural saline (2 pg/ml ± 2), while other hormones showed no significant differences. The authors conclude that high epidural anesthesia in awake unsedated dogs: 1) almost completely abolishes the normal cardiovascular response to hypoxemia while promoting vasopressin secretion; 2) preserves the ventilatory response to hypoxemia; and 3) is associated with increased vasopressin concentrations, most likely to compensate for decreased cardiac filling and/or arterial blood pressure when sympathoadrenal responses are impaired. Thus, the changes in cardiovascular vital signs in response to severe hypoxemia are markedly blunted when spinal sympathetic outflow is selectively eliminated by epidural anesthesia.