Drought events are complex and uncertain, and drought sensitivity assessments are beneficial for ecological conservation and improving early warning mechanisms. Understanding ecosystems' response to drought is vital for ecological conservation and climate change research. Ecological drought, a concept describing drought from an ecosystem perspective, lacks sufficient research regarding its effects on agricultural ecosystems, especially in the Huang-Huai-Hai Plain. This study identifies ecological drought using a three-dimensional clustering method and examines its spatiotemporal patterns. By developing the Winter Wheat Ecological Drought Sensitivity Index (EDSIWW), it quantifies the impact of drought on winter wheat and employs machine learning models to assess environmental factors influencing this relationship. Findings show that 26 ecological drought events occurred between 2000 and 2022, predominantly in autumn and winter, with an average duration of 132 days, affecting over 80% of the study area. These events coincided with winter wheat's growth cycle, significantly impacting its gross primary production, followed by solar-induced chlorophyll fluorescence, leaf area index, and near-infrared reflectance. Environmental factors like land surface temperature, soil moisture, and vapor pressure deficit exhibited complex influences on wheat's sensitivity to drought. This research offers new insights into the drought-wheat relationship and provides scientific guidance for agricultural management and drought prevention.