Abandoned cropland (AC) is one of the important types of land cover and land use change that poses a threat to food security, nature conservation, and ecosystem protection. Despite its significance, detailed spatial distribution patterns of AC across expansive geographic regions remain under-explored. Therefore, this study 1) proposes a novel framework designed to precisely determine the timing and extent of AC, and evaluates its effectiveness in the Beijing-Tianjin-Hebei region from 2000 to 2020; 2) analyses the spatial and temporal evolution of AC; 3) investigates the relationship between AC and both physical geographic and socio-economic factors. The proposed AC recognition framework integrates the method of change detection. Initially, regions changing are pinpointed, followed by categorizing pre- and post-change images to ascertain the magnitude and timing of these shifts, enabling the pinpointing of AC locations and their temporal onset. The results showed that the accuracy of land cover classification mapping ranged from 80.83% to 99.08%. Specifically, the user's accuracy of the proposed method varied between 79.29% and 89.38%. We discerned that AC was mainly distributed in the northwest water conservation belt, adopting an inverted “L” pattern. Additionally, physical geographic factors predominantly influenced AC. The use of remote sensing plays a crucial role in assessing AC, which is of great significance for understanding the evolution of abandoned landscapes and analyzing the relationship between abandonment and geology, environment, and socio-economy.