The cloud effect correction is very significant to satellite remote sensing of atmospheric trace gases from near-infrared spectrum, even with a small amount of cloud, the remaining effects can still cause retrieval error even reaching up to 100%.Atmospheric carbon monoxide (CO) is one of the most important pollutants in the troposphere. This study, takeing CO as the example, described the cloud effect and the cloud correction on retrieval result from SCIAMACHY measurement. To validate the correction method, we compared both corrected and uncorrected results with the independent ground based FTIR measurements. After the correction, the agreement between satellite observations and FTIR measurements further improves.