Landfill methane (CH4) emissions account for ~ 10% of all anthropogenic CH4 emissions globally, amounting to ~ 50 Tg/year. Contrasted by “top-down” atmospheric inversion results, the mainstream “bottom-up” emission inventories, which use the Intergovernmental Panel on Climate Change’s (IPCC) model, exhibit significant bias due to inaccurate a priori decay constant (k) estimations. We improved the k estimation method by incorporating composition- and environment-specific corrections, which are readily integrated into the IPCC’s model. We demonstrate that the accuracies of CH4 emission predictions are significantly improved by using the corrected k values, which are benchmarked against the atmospheric inversion results. We extend the emission estimations to landfills worldwide and reveal up to 209% underestimation in individual landfills and several tens of Tg/year of potentially overlooked CH4 emission globally. Our findings highlight the importance of prioritizing landfill CH4 emission monitoring and reduction as one of the most cost-effective mitigation options to achieve current climate goals.