Background: Peripheral venous blood gas (pVBG) analysis in replacement of arterial blood gas (ABG) is limited by the unpredictable differences between arterial and venous values, especially for PCO2 and pH (Delta PCO2 and Delta pH). Objectives: We hypothesized that, using the theoretical relationship linking SvO(2) and blood flow, we could diminish the effect of local circulatory conditions on Delta PCO2 and Delta pH and thereby increase pVBG validity. Methods: This was a prospective cross-sectional study performed in emergency patients requiring a blood gas analysis in which ABG and pVBG were performed simultaneously. The data of 50 randomly selected patients (model group) were used for developing two equations to correct PvCO(2) and pHv according to the peripheral SvO(2) (SpvO(2)) level. The formulas derived were PvCO(2cor) = PvCO(2) - 0.30 x (75 - SpvO(2)), and pHv(cor) = pHv + 0.001 x (75 - SpvO(2)). The validity of the corrected values was then tested on the remaining population (validation group). Results: There were 281 patients included in the study, mainly for dyspnea. Delta PCO2 and Delta pH were strongly correlated with SpvO(2) (r(2) = 0.62 and r(2) = 0.53, respectively, p < 0.001). Using the data of the model group, we developed equations that we applied on the validation group. We found that the corrected values were more valid than the raw values for detecting a PaCO2 > 45 mm Hg (AUC ROC = 0.96 +/- 0.01 vs. 0.89 +/- 0.02, p < 0.001), a PaCO2 < 35 mm Hg (AUC = 0.95 +/- 0.02 vs. 0.84 +/- 0.03, p < 0.001), a pHa < 7.35 (AUC = 0.97 +/- 0.01 vs. 0.95 +/- 0.02, p < 0.05), or a pHa > 7.45 (AUC = 0.91 +/- 0.02 vs. 0.81 +/- 0.04, p < 0.001). Conclusions: The variability of Delta PCO2 and Delta pH is significantly lowered when the venous values are corrected according to the SpvO(2) value, and pVBG is therefore more accurate and valid for detecting an arterial abnormality. (C) 2013 Elsevier Inc.