Although the concept of ecological drought was first defined by the Science for Nature and People Partnership (SNAPP) in 2016, there remains no widely accepted ecological drought monitoring index. Therefore, this study constructed a new ecological drought index, the standardized ecological water deficit index (SEWDI). The SEWDI is based on the difference between ecological water requirement and consumption, referred to as the standardized precipitation index (SPI) method, which was used to monitor ecological drought in Northwestern China (NWRC). The performance of the SEWDI was compared with that of other widely–used drought indices, including standardized root soil moisture index (SSI), self–calibrated Palmer drought index (scPDSI), standardized precipitation–evaporation drought index (SPEI), and SPI, using the Pearson correlations between these indices and standardized normalized difference vegetation index (SNDVI) under different time scales, wetness and water use efficiencies (WUE) of vegetation. The SEWDI at a 12–month scale was decomposed in NWRC during 1982–2015 using the rotational empirical orthogonal function (REOF) in order to demarcate five ecological drought regions, including southeastern sub–region (SE), southwestern sub–region (SW), northwestern sub–region (NW), northeastern sub–region (NE), and central sub–region (CT). The characteristics of ecological drought in NWRC, such as intensity, duration, and frequency, were extracted using the run theory. The return periods of five types of drought were calculated using wavelet analysis. Results showed that the performance of SEWDI in monitoring ecological drought was the best among the drought indices under different time scales, and the 12–month–scale was largely unaffected by wetness and WUE. Results of monitoring indicated that serious ecological droughts in the NWRC mainly occurred in 1982–1986, 1990–1996, and 2005–2010, primarily in SE, SW, and CT, SW and NE, and NW, NE, and CT, respectively. Furthermore, SEWDI in the arid SW, NW and CT showed a longer return period compared to that in the humid NE and SE, and the evolution of ecological drought in arid regions was mainly influenced by meteorological drought and the scarcity of root soil moisture. This study provides ant approach for quantifying ecological drought severity across natural vegetation areas which can be employed by decision makers