We determined the clinical efficacy of using the venous CO2 value, as obtained with routine "electrolytes", in acid-base management. Venipuncture samples for venous CO2 content and chloride concentrations were obtained in 336 patients with arterial blood pH, PaO2, PaCO2, and oxygen saturation determinations. The linear correlation between actual calculated arterial HCO3- and the measured venous CO2 was significant (P less than .001). Using venous CO2, chloride, and arterial pH values, we present various prediction equations for estimating arterial HCO3-. We determined the effects of time delay, exposure to air, and acute changes in arterial blood pH and PaCO2 on venous CO2 levels. Venous CO2 determinations should not be substituted for the arterial HCO3 value in the Henderson-Hasselbalch equation to calculate arterial blood pH or PaCO2. Clinically, the venous CO2 value has little direct use, but when venous CO2 content is abnormal, it should alert the clinician to the need for obtaining arterial blood gas and pH values.