In order to gain a deeper understanding of the characteristics of ground-dwelling spider communities, the ecological niche characteristics of the main populations, and their relationship with environmental factors in the loess hilly area of China, this study selected typical vegetation communities (Pinus tabuliformis forest, Prunus sibirica forest, Hippophae rhamnoides forest, Robinia pseudoacacia forest, Populus simonii forest and farmland) formed after the conversion of farmland to forest (grassland) in the loess hilly area as the research site. From August to November 2022, the trap method was used to collect ground-dwelling spiders in each site, and the collected spiders were morphological identified. The results showed that: (1) A total of 2463 ground-dwelling spider specimens were collected, belonging to 18 families, 36 genera, and 49 species, including 13 main species. Returning farmland to forest has had a significant impact on the evenness index and dominance index of ground-dwelling spiders (P < 0.05), but had no significant impact on the richness index and diversity index. The highest evenness index of ground-dwelling spiders was found in Prunus sibirica forest and farmland, with a evenness index of 0.82. The lowest evenness index was in the Populus simonii forest, with a evenness index of 0.49. The highest dominance index of ground-dwelling spiders was Prunus sibirica forest, with a dominance index of 7.38, while the lowest dominance index was Populus simonii forest, with a dominance index of 2.13. The Jaccard similarity index between most vegetation in the study area ranged from 0.4 to 0.5, and the overall composition of ground-dwelling spider communities showed moderate dissimilarity. (2) The largest ecological niche width of the main ground-dwelling spider was the Xysticus ephippiafus, with an ecological niche width of 5.30; The Pardosa schenkeli had the smallest niche width, with a niche width of 1.54. The distribution range of ecological niche overlap values of the main ground-dwelling spiders was [0.06, 1], with high niche overlap accounting for 23.08% of the total logarithm, moderate overlap accounting for 28.21% of the total logarithm, and low niche overlap accounting for 48.72% of the total logarithm. The sum of the ecological response rates of the 13 main ground-dwelling spiders in the study area was negative, indicating that the main ground-dwelling spiders were in a declining stage. (3) The redundancy analysis (RDA) results showed that herbaceous coverage, litter thickness, available potassium, electrical conductivity, and organic matter were key driving factors affecting the distribution and niche characteristics of major ground-dwelling spiders. Therefore, this study indicated that the cultivation of artificial forests is beneficial for the restoration and protection of local ground-dwelling spider diversity. The main ground-dwelling spiders in the area of returning farmland to forest, as well as the main species and the environment, have not yet reached a relative dynamic balance.