Introduction: Quantification of physiological dead space (VDphys) provides important insight into the efficiency of ventilation and its relation to pulmonary perfusion. Arterial to end-tidal carbon dioxide difference [P(a-ET)CO₂] may provide a valuable surrogate measure of VDphys and may be useful as an index of weaning success. Aim: To evaluate the utility of [P(a-ET)CO₂] as a weaning index. Methods: This prospective study enrolled 52 invasively mechanically ventilated adult patients treated in Medical Intensive Care Unit (MICU) between December 2010 and December 2011. The end-tidal carbon dioxide concentration was measured using a side stream capnograph at each attempt at decreasing ventilatory support and when they were ready to be weaned. A receiver operating characteristic (ROC) curve was constructed for weaning success, progressive weaning and extubation success. A cut-off point was obtained from these curves from which the sensitivity, specificity, positive and negative predictive values were obtained. Results: The ability of [P(a-ET)CO₂] as a predictor of progressive reduction in ventilator support (n = 118), predictor of weaning from mechanical ventilation (n = 40) and for extubation success (n = 39) was evaluated. The area under the curve (AUC) for progressive weaning, spontaneous breathing trial and extubation success were 0.852, 0.905 and 0.702 and a threshold of 10.5 mm Hg, 9.4 mm Hg and 9.5 mm Hg respectively were obtained. Conclusion: P(a-ET)CO₂ of ≤ 10 mm Hg may be used as an index of weaning during progressive weaning from mechanical ventilation, spontaneous breathing trial and to predict success of extubation.