The three "carbon gases" carbon monoxide (CO), methane (CH 4) and carbon dioxide (CO 2) are important atmospheric constituents affecting air quality and climate. The near-infrared nadir spectra of reflected solar radiation measured by SCIAMACHY on-board ENVISAT contain information on the vertical columns of these gases which we retrieve using the scientific algorithm WFM-DOAS. For CH 4 and CO 2 our main data products are dry air column averaged mixing ratios (XCH 4 in ppmv and XCO 2 in ppmv) determined by simultaneous measurements of the dry air mass obtained from oxygen (O 2) for CO 2 and CO 2 for methane. Our CO data product is the CO vertical column in molecules/cm 2. The SCIAMACHY data set is unique because of the high sensitivity of the near-infrared measurements with respect to concentration changes in the atmospheric boundary layer. This sensitivity is a pre-requisite to get information on regional surface sources and sinks which are currently only poorly constrained globally by atmospheric measurements. We present a short overview about the latest versions of the retrieval algorithm (WFM-DOAS version 0.5 for CO and CH 4 and version 0.4 for CO 2) and processed data (year 2003 data set based on Level 1 version 4 spectra) including comparisons with global and local reference data (MOPITT, ground based FTS, and global models).