Methane (CH4) is an important climate forcer, contributing about 17% of the total radiative forcing by long living greenhouse gases. The Po basin is one of the most polluted and densely populated areas in Europe representing an important source region for CH4. The aim of this work was to test an inter-species correlation approach to derive estimates of anthropogenic CH4 emissions for the period 2015–2019 from the Po basin by combining CO bottom-up inventory data and continuous CH4 and CO observations from a mountain site in the northern Italy. The tested methodology suggested lower emissions in respect to EDGAR (−17%) and the Italian National Inventory (−40%) for the Po basin. However, despite the two bottom-up inventories, the emissions derived from the atmospheric observations reported an increasing tendency from 2015 to 2019 for the CH4 emissions. A sensitivity study revealed that using different subsets of the atmospheric observations implied a difference of 26% in the CH4 emission estimates. The highest agreement with two bottom-up CH4 inventories (EDGAR and the Italian national inventory) were obtained when atmospheric data were strictly selected for periods representative of air mass transport from the Po basin. Our study identified various challenges when using this methodology as a benchmark to verify bottom-up CH4 inventories. Issues could be attributed to the annual aggregation of the proxies used to derive the emission amounts, to the CO bottom-up inventory used as input information and to the relatively high sensitivity of the results to the different subsets of the atmospheric observations. However, the use of different bottom-up inventories as input data for CO emissions can potentially provide information that should be carefully considered for the purpose of integrating CH4 bottom-up inventories. •We tested and used an inter-species correlation approach to calculate anthropogenic CH4 emissions from the Po basin.•Atmospheric observations from a mountain site were used.•Differences with bottom-up inventories were pointed out for 4 investigated years.•An accurate selection of the atmospheric data is pivotal for getting reliable information.•Yearly aggregation of data and input CO emissions led to further uncertainties.