First results concerning the retrieval of tropospheric carbon monoxide (CO) from satellite solar backscatter radiance measurements in the near-infrared spectral region (~2.3µm) are presented. The Weighting Function Modified (WFM) DOAS retrieval algorithm has been used to retrieve vertical columns of CO from SCIAMACHY/ENVISAT nadir spectra. We present detailed results for three days from the time periode January to October 2003 selected to have good overlap with the daytime CO measurements of MOPITT onboard EOS Terra. Because the WFM-DOAS Version 0.4 CO columns presented in this paper are scaled by a constant factor of 0.5 to compensate for an obvious overestimation we focus on the variability of the retrieved columns rather than on their absolute values. It is shown that plumes of CO resulting from, e.g. biomass burning in Africa, are detectable with single overpass SCIAMACHY data. Globally, the SCIAMACHY CO columns are in reasonable agreement with the Version 3 CO column data product of MOPITT. For example, for measurements over land, where the quality of the data is typically better than over ocean due to higher surface reflectivity, the standard deviation of the difference with respect to MOPITT is in the range 0.4-0.6x10¹⁸ molecules/cm² and the linear correlation coefficient is between 0.4 and 0.7. The level of agreement between the data of both sensors depends on time and location but is typically within 30% for most latitudes. In the southern hemisphere outside Antarctica SCIAMACHY tends to give systematically higher values than MOPITT. More studies are needed to find out what the reasons for the observed differences with respect to MOPITT are and how the algorithm can be modified to improve the quality of the CO columns as retrieved from SCIAMACHY.