The toad-headed lizard (Phrynocephalus guinanensis) exclusively distributes in an area of sand dunes with approximately 20 km wide and 30 km long in Guinan County of Qinghai Province. While the above sand dunes were geographically isolated by surrounded grasslands, relatively larger number of active lizards were encountered in sand dunes than in grassland in the contact area of sand and grass habitats. Three line transects with 150 traps were used to study the spatial distribution of active caves and the potentially available insect resources for P. guinanensis in grassland and sand dunes in both sides of the contact area. The relationship between cave distribution and insect resource availability was analyzed. We successfully captured 1236 individual insects belonging to 2 classes, 10 orders, 26 families, and 42 species. Among them, 978 Insecta individuals were observed, comprising 7 orders, 21 families, and 37 species, while the remaining 258 Arachnoidea individuals comprised 3 orders, 5 families, and 5 species. We also discovered that among the 570 individuals captured in the grassland, 2 classes, 9 orders, 22 families, and 37 species were present, whereas 666 individuals were captured in sand dunes and included 2 classes, 9 orders, 21 families, and 33 species. The Shannon-Wiener diversity index, Simpson diversity index, and an evenness index for grassland insects were all larger than the corresponding indexes for the insects in sand dunes, which indicated that the insect resources were more abundant and homogeneous in grasslands. Our results also showed that more active caves existed in sand dunes than in grasslands, and their quantity in sand dunes increased with a decrease in the distance to the grassland-sand dune boundary; however, no correlation was found between the quantity of active caves in grasslands and their distance to the boundary, which supports the idea that a homogeneous spatial distribution of caves exists in grasslands. Moreover, we found no correlation between the distance to the boundary, and the number of insect species and individuals in grassland and the number of insect individuals in sand dunes; however, a significantly negative correlation was obtained between the distance to the boundary and the number of insect species in sand dunes. Therefore, our results concluded that a close relationship exists between the spatial distribution of active caves of P. guinanensis and the distribution of insect species in sand dunes.