Evapotranspiration (ET) plays an important role in exchange of water budget and carbon cycles over the Inner Mongolia autonomous region of China (IMARC). However, the spatial and decadal variations in terrestrial ET and drought over the IMARC in the past was calculated by only using sparse meteorological point-based data which remain quite uncertain. In this study, by combining satellite and meteorology datasets, a satellite-based semi-empirical Penman ET (SEMI-PM) algorithm is used to estimate regional ET and evaporative wet index (EWI) calculated by the ratio of ET and potential ET (PET) over the IMARC. Validation result shows that the square of the correlation coefficients (R 2) for the four sites varies from 0.45 to 0.84 and the root-mean-square error (RMSE) is <0.78 mm. We found that the ET has decreased on an average of 4.8 mm per decade (p = 0.10) over the entire IMARC during 1982-2009 and the EWI has decreased on an average of 1.1% per decade (p = 0.08) during the study period. Importantly, the patterns of monthly EWI anomalies have a good spatial and temporal correlation with the Palmer Drought Severity Index (PDSI) anomalies from 1982 to 2009, indicating EWI can be used to monitor regional surface drought with high spatial resolution. In high-latitude ecosystems of northeast region of the IMARC, both air temperature (T a) and incident solar radiation (R s) are the most important parameters in determining ET. However, in semiarid and arid areas of the central and southwest regions of the IMARC, both relative humidity (RH) and normalized difference vegetation index (NDVI) are the most important factors controlling annual variation of ET.