<p>The COVID-19 pandemic resulted in reduced anthropogenic carbon dioxide (CO<sub>2</sub>) emissions during 2020 in large parts of the world. We report results from a first attempt to determine whether a regional-scale reduction of anthropogenic CO<sub>2</sub> emissions during the COVID-19 pandemic can be detected using space-based observations of atmospheric CO<sub>2</sub>. For this purpose, we have analysed a small ensemble of satellite retrievals of column-averaged dry-air mole fractions of CO<sub>2</sub>, i.e. XCO<sub>2</sub>. We focus on East China because COVID-19 related CO<sub>2</sub> emission reductions are expected to be largest there early in the pandemic. We analysed four XCO<sub>2</sub> data products from the satellites Orbiting Carbon Observatory-2 (OCO-2) and Greenhouse gases Observing SATellite (GOSAT). We use a data-driven approach that does not rely on <i>a priori</i> information about CO<sub>2</sub> sources and sinks and ignores atmospheric transport. Our approach utilises the computation of XCO<sub>2</sub> anomalies, ΔXCO<sub>2</sub>, from the satellite Level 2 data products using a method called DAM (Daily Anomalies via (latitude band) Medians). DAM removes large-scale, daily XCO<sub>2</sub> background variations, yielding XCO<sub>2</sub> anomalies that correlate with the location of major CO<sub>2</sub> source regions such as East China. We analysed satellite data between January 2015 and May 2020 and compared monthly XCO<sub>2</sub> anomalies in 2020 with corresponding monthly XCO<sub>2</sub> anomalies of previous years. In order to link the XCO<sub>2</sub> anomalies to East China fossil fuel (FF) emissions, we used XCO<sub>2</sub> and corresponding FF emissions from NOAA’s (National Oceanic and Atmospheric Administration) CarbonTracker version CT2019 from 2015 to 2018. Using this CT2019 data set, we found that the relationship between target region ΔXCO<sub>2</sub> and the FF emissions of the target region is approximately linear and we quantified slope and offset via a linear fit. We use the empirically obtained linear equation to compute ΔXCO<sub>2</sub><sup>FF</sup>, an estimate of the target region FF emissions, from the satellite-derived XCO<sub>2</sub> anomalies, ΔXCO<sub>2</sub>. We focus on October to May periods to minimize contributions from biospheric carbon fluxes and quantified the error of our FF estimation method for this period by applying it to CT2019. We found that the difference of the retrieved FF emissions and the CT2019 FF emissions in terms of the root-mean-square-error (RMSE) is 0.39 GtCO<sub>2</sub>/year (4 %). We applied our method to NASA’s (National Aeronautics and Space Administration) OCO-2 XCO<sub>2</sub> data product (version 10r) and to three GOSAT products. We focus on estimates of the relative change of East China monthly emissions in 2020 relative to previous months. Our results show considerable month-to-month variability (especially for the GOSAT products) and significant differences across the ensemble of satellite data products analysed. The ensemble mean indicates emission reductions by approximately 8 % ± 10 % in March 2020 and 10 % ± 10 % in April 2020 (uncertainties are 1-sigma) and somewhat lower reductions for the other months in 2020. Using only the OCO-2 data product, we obtain smaller reductions of 1–2 % (depending on month) with an uncertainty of ± 2 %. The large uncertainty and the differences of the results obtained for the individual ensemble members indicates that it is challenging to reliably detect and to accurately quantify the emission reduction. There are several reasons for this including the weak signal (the expected regional XCO<sub>2</sub> reduction is only on the order of 0.1–0.2 ppm), the sparseness of the satellite data, remaining biases and limitations of our relatively simple data-driven analysis approach. Inferring COVID-19 related information on regional-scale CO<sub>2</sub> emissions using current satellite XCO<sub>2</sub> retrievals likely requires, if at all possible, a more sophisticated analysis method including detailed transport modelling and considering <i>a priori</i> information on anthropogenic and natural CO<sub>2</sub> surface fluxes.</p>