Land use change forced by artificial-fixed sandy vegetation recovery in an arid area has influenced the community and diversity of ground-dwelling arthropods. Up to now, few investigators have focused on responses of ground spider and beetle communities to artificial fixed-sand vegetation reconstruction, and its key influencing elements. In this study, two shrubs planted types (Haloxylon ammodendron plantations, HAP and Tamarix ramosissma plantations, TRP) without irrigation and natural shrub forest (natural shrub forest, NSF) selected for comparison with the shrub planted habitat as control. Ground spider and beetle communities in the three habitats collected by pitfall trapping, and environmental aspects related to the scattering of spiders and beetles were explored. One-way analysis was used to compare the difference of the activity density, taxa richness and Shannon-Wiener diversity of ground spiders and beetles as well as soil and vegetation factors among the three habitats. Further, analysis of variance and multivariate analysis was used to show the difference of ground spiders and beetles among three habitats. Redundancy analysis (RDA) and Canonical correlation analysis (CCA) were performed to quantify the relative contribution of the vegetation and soil ecological factors to the variation in the ground spider and beetle assembly and to identify key drivers of changes in these compositions. The results showed that significant differences in the diversity and the composition of the community on ground spiders and beetles were observed among the three habitats. Further, the activity density of ground spiders and the Shannon-wiener index of ground beetles in HAP and TRP habitats was greater than that in NSF habitats, a contrary pattern on the activity density of ground beetles and the Shannon-wiener index of ground spiders was observed. There are likewise some differences in the composition of spider and beetle communities between HAP and TRP habitats. The activity density and Shannon-wiener index of ground spiders and species richness of ground beetles in TRP habitats were higher than those in HAP habitats. The further analysis we discovered that Lycosidae, Gnaphosidae, Linyphiinae, Theridiidae, and some beetle species belonged to Tenebrionidae, Carabidae, Curculionidae were preferred for NSF or HAP or TRP habitats, which determined the assemblage of ground spiders and beetles. The consequences of pRDA show that herbaceous biomass, litter production, soil sand content, soil electrical conductivity and shrub cover were key environmental factors affecting the distribution of ground spiders, which contributed to 82.1% of the variation in the ground spiders. And the consequences of pCCA show that shrub cover, herbaceous biomass, pH and soil sand content were fundamental environmental factors affecting the distribution of ground beetles, which contributed to 60.6% of the variation in the ground beetles. In short, the restoration of artificial sand-fixing shrubs drives the change of vegetation and soil environment, which determined the distribution pattern of ground spiders and beetles in the desert-oasis transition zone.