Abstract: Rural settlements of people are normally formed based on the geography, blood, and agricultural production. Therefore, the evolution of rural settlements can be used to clarify the special human-land relationship in rural areas, while the spatial correlation between rural settlements and geographical factors can also provide important information for the subsequent scheme of plan and reconstruction. However, there are different features in spatiotemporal differentiation and correlation in different geographical scale during the evolution of rural settlements. Most previous studies usually focused on the evolution features of rural settlements in single medium-micro scale. The multi-scale analysis on evolution of rural settlements and spatial correlation is lacking, particularly on systematical understanding the behind mechanisms. This study aims to investigate the evolution characteristics of rural settlements in Jiangsu Province during 2009 to 2018, using the kernel density analysis, landscape fractal dimension index, and hotspots analysis. The kernel density was used to identify the density change of rural settlements at the provincial level, while the landscape fractal dimension index was used to analyze the shape change of rural settlements at the county level. The area was divided into the grid of 10 km×10 km, in order to explore the dynamic change of rural settlements. The loss and gain of rural settlements were presented at the patch level and grid level. Additionally, the relationships between the changes of rural settlements and roads, water bodies, cities, towns, and relief amplitude, were analyzed, and then measured based on spatial access. Getis-Ord Gi* algorithm was used to identify the hotspots of rural settlement change. The results showed that there was a growth rate of 1.35% for the total area of rural settlements, while there was a total loss of rural population, over ten million, indicating an inconsistent condition in the human-land population from 2009 to 2018 in Jiangsu Province. There was a space change in a typical feature with large-scale loss and concentration gain. The density variation in the loss and gain of rural settlements presented a concentration change along the Yangtze River and sea coast. The shapes of rural settlements were in an irregular distribution, particularly on the significant difference in space. The regularity of rural settlements in the southern Jiangsu was better than that in central and northern Jiangsu. Most changes of rural settlements occurred within the scope of over 60 minutes away from the road, water body, city and town. However, it was not observed an obvious spatial access orientation of geographical elements. The changes of rural settlements randomly occurred in the way of self-organization from farmers in the majority counties. The typical hotspots from loss and gain of rural settlements located in Nanjing, while the gain hotspots can also be observed in Nantong. This finding can contribute to understanding the spatiotemporal relationship and farmers' decision-making behavior that involved in the process of rural settlements evolution under socioeconomic transformation period, and thereby to promote regional spatial optimization and reconstruction of rural settlements.