The energy structure in China is facing sustainable transformation and development. Hydropower, as a source of clean energy, is widely built up in the upper reaches of Changjiang River drainage basin where great terrain differences occurred. However, there is a lack of research on the greenhouse gases (GHGs) emissions of reservoirs in the hydropower life cycle in a specific region, and the inconsistent estimation methods lead to poor comparability of the estimation results. Based on Tier1 in the IPCC National Greenhouse Gas Inventories, this study explores the life cycle GHGs emissions of 24 medium and large reservoirs in the upper reaches of Changjiang River, and conducts uncertainty and sensitivity analysis of model parameters through Monte Carlo simulation. The results show that the average carbon emissions of the 24 reservoirs are distributed in the range of 0.0342-140.59 Tg CO_2eq. The total carbon emissions reached 264.05 Tg CO_2eq (CO₂ emissions accounted for 9.12%, CH₄ emissions accounted for 90.88%). The average carbon emission per unit power generation is 3.30 g CO_2eq/(kW·h) (0.01-17.64 g CO_2eq/(kW·h)), and the minimum and maximum values appeared in Jinping Ⅱ Reservoir and Pengshui Reservoir, respectively. In addition, sensitivity analysis found that among the parameters involved in the model, the nutritional adjustment coefficient played a leading role in the uncertainty of the final estimation results, while the ratio of total downstream flux of CH₄ to the flux of CH₄ from a reservoir's surface to the atmosphere had the weakest sensitivity to results. This provides good guidance to select parameters when using this model and to carry out sampling detection in the future.