Objective:To describe global hemodynamics and splanchnic perfusion changes in response to acute modifications in PaCO2in hemodynamically stable patients.Design:Prospective, randomized crossover study.Setting:Medical-surgical intensive care unit at a community hospital (400,000 inhabitants).Patients:Ten critically III patients who were sedated, paralyzed, and mechanically ventilated.Interventions:Hypercapnia and hypocapnia were obtained by increasing and reducing instrumental deadspace in random order. After each intervention, patients returned to the basal condition. Each period lasted 80 min: 20 min to achieve stable PaCO2and 60 min for tonometer equilibration. In each period, global hemodynamic variables and tonometric data were collected. The periods were compared using analysis of variance.Measurements and Main Results:Acute hypercapnia (PaCO2from 40 ± 3 to 52 ± 3 torr,p< .05) increased cardiac index (3.43 ± 0.37 vs. 3.97 ± 0.43 mL/min/m2,p< .05), heart rate (95 ± 6 vs. 105 ± 3 beats/min,p< .05), and mean pulmonary artery pressure (21 ± 1 vs. 24 ± 1 mm Hg,p< .05) and reduced systemic vascular resistance (992 ± 98 vs. 813 ± 93 dyne-sec/cm5,p< .05) and oxygen extraction ratio (27 ± 3% vs. 22 ± 2%,p< .05). Standardized intramucosal PCO2increased from 49 ± 2 to 61 ± 3 torr (p< .05) with an associated decrease in calculated intramucosal pH ([pHi] 7.35 ± 0.03 vs. 7.25 ± 0.02,p< .05), but the gastro-arterial PCO2gradient (ΔPCO2) did not change. Acute hypocapnia (PaCO2from 41 ± 3 to 34 ± 3 torr,p< .05; pH 7.41 ± 0.01 to 7.47 ± 0.02,p< .05) induced slight increments in systemic vascular resistance (995 ± 117 vs. 1088 ± 160 dyne·sec/cm5,p< .05) and oxygen extraction ratio (28 ± 2% vs. 30 ± 2%,p< .05). Standardized intramucosal PCO2decreased (50 ± 4 vs. 44 ± 3 torr,p< .05), pHi increased (7.33 ± 0.03 vs. 7.36 ± 0.02;p< .05), but ΔPCO2did not change.Conclusions:In this small group of stable patients, moderate acute variations in PaCO2had a significant effect on global hemodynamics, but splanchnic perfusion, assessed by ΔPCO2, did not change. In these conditions, the use of pHi to evaluate gastric perfusion appears unreliable.