Résumé : Oxygenated volatile organic compounds (OVOCs) have a significant impact on tropospheric chemistry. Large uncertainties are still associated with their atmospheric budget estimates, primarily due to the paucity of global and consistent measurements to constrain the distribution and magnitude of their emissions and those of their precursors. The spatially dense observations from the nadir-viewing Infrared Atmospheric Sounding Interferometer (IASI), on board the Metop platforms, can help reducing these uncertainties by providing trace gas quantification throughout the globe. However, the weak and sometimes broadband spectral signatures of the OVOCs in the thermal infrared make their retrieval particularly challenging.We have applied the ANNI (Artificial Neural Network for IASI) general retrieval framework version 3 to derive total columns of five major OVOCs, namely methanol, formic and acetic acid, PAN and acetone, from the IASI observations. This method, which we briefly explain, reveals particular sensitivity to the detection of OVOCs and allows the production in near-real time of daily global distributions of these compounds. Furthermore, the ANNI computing efficiency permits the processing of the entire IASI/Metop-A and -B observational time series. We present and describe the decadal climatology of each species, focusing on the contribution from terrestrial vegetation and biomass burning, and on their seasonal and interannual variabilities. Moreover, the retrieved IASI data allow the production of daily regional snapshots, which we use to illustrate local OVOC enhancements (e.g., in fire plumes) detected in various environments. Finally, the OVOC abundance ratios to IASI carbon monoxide and their chemistry in fire plumes are discussed.