OBJECTIVE One-lung ventilation (OLV) may be complicated by hypoxemia. Ventilatory efficiency, defined as the ratio of minute ventilation to carbon dioxide output (VE/VCO2), is increased with ventilation/perfusion mismatch and pulmonary artery hypertension, both of which may be associated with hypoxemia. Hence, the authors hypothesized increased VE/VCO2 will predict hypoxemia during OLV. METHODS Prospective observational study. METHODS Single-center, university, tertiary care hospital. METHODS The study comprised 50 consecutive lung resection candidates. METHODS All patients underwent cardiopulmonary exercise testing before surgery. Patients who required inspired oxygen fraction (FiO2) ≥0.7 to maintain arterial oxygen (O2) saturation >90% after 30 minutes of OLV were considered to be hypoxemic. The Student t or Mann-Whitney U test were used for comparison of patients who became hypoxemic and those who did not. Multiple regression analysis adjusted for age, sex, and body mass index was used to evaluate which parameters were associated with the VE/VCO2 slope. Data are summarized as mean ± standard deviation. RESULTS Twenty-four patients (48%) developed hypoxemia. There was no significant difference in age, sex, and body mass index between hypoxemic and nonhypoxemic patients. However, patients with hypoxemia had a significantly higher VE/VCO2 slope (30 ± 5 v 27 ± 4; p = 0.04) with exercise and lower partial pressure of oxygen/FiO2 (129 ± 92 v 168 ± 88; p = 0.01), higher mean positive end-expiratory pressure (6.6 ± 1.5 v 5.6 ± 0.9 cmH2O; p = 0.02), and lower mean pulse oximetry O2 saturation/FiO2 index (127 ± 20 v 174 ± 17; p < 0.01) during OLV. Multiple regression showed VE/VCO2 to be independently associated with the mean pulse oximetry O2 saturation/FiO2 index (b = -0.28; F = 3.1; p = 0.05). CONCLUSIONS An increased VE/VCO2 slope may predict hypoxemia development in patients who undergo OLV.