Carbon forestry mitigation potential estimates at the global level are limited by the absence or simplicity of national level estimates, and similarly national-level estimates are limited by absence of regional-level estimates. The present study aims to estimate the mitigation potential for a large diverse country such as India, based on the GTAP global land classification system of agro-ecological zones (AEZs), as well the Indian AEZ system. The mitigation potential for short and long rotation plantations and natural regeneration was estimated using the GCOMAP global forest model for one scenario that included only wastelands (29 Mha), and the second one that included wastelands plus long fallow and marginal croplands (54 Mha). Under the $100 carbon price case, significant additional area (3.6 Mha under the waste land scenario and 6.4 Mha under the enhanced area scenario) and carbon mitigation is gained in the short-term (2025) compared to the baseline when using the GTAP land classification system. However, 35% additional C-stock gain is achieved in the second scenario using the Indian AEZ system. This difference highlights the importance of the impact of a land classification system on aggregate mitigation potential estimates. Uncertainty involved in the estimates of national level mitigation potential needs to be reduced, by generating reliable estimates of carbon stock gain and losses, and cost and benefit data, for land use sector mitigation options at a scale disaggregated enough to be relevant for national mitigation planning.