[Objective] The application prospects of using landslide deposits as a new type of soil improvement material were explored, and the changes in soil nutrients after landslide deposit improvement were analysed in order to provide new approaches for local landslide deposit utilisation and purple soil rehabilitation. [Methods] Purple soil improved by landslide deposits was selected as the research object, and the quality of the improved soil was analysed using weighted grey correlation analysis. [Results] At the same fertility irrigation level, adding collapsed deposits significantly increased the nutrient content of topsoil (0—20 cm) compared to the blank control group (Y₀). The average total nitrogen content in this layer of soil increased by 3.26% to 10.60%, the average total phosphorus content increased by 10.61% to 43.33%, the average organic matter content increased by 9.14% to 24.12%, the average alkaline nitrogen content increased by 5.76% to 12.73%, the average effective phosphorus content increased by 24.75% to 46.22%, the average available potassium content increased by 16.21% to 28.96%, and the soil electrical conductivity (EC) increased by 11.32% to 27.41%. The different treatment schemes had no significant effect on soil pH value. Using the AHP grey correlation analysis evaluation model, a preliminary evaluation of the soil quality of each treatment group was conducted. Within the topsoil (0—20 cm), the grey correlation degree of each treatment group was ranked as follows: Y₂ 〔rock soil composite ratio (RSCR): 20%〕 > Y₃ (RSCR: 30%) > Y₄ (RSCR: 40%) > Y₁ (RSCR: 10%) > Y₀ (RSCR: 0%) > Y₅ (RSCR: 50%). The grey correlation degree of each treatment group within the deep soil range of 20—40 cm was ranked as follows: Y₅ > Y₄ > Y₃ > Y₂ > Y₁ > Y₀. [Conclusion] An appropriate proportion of composite treatment between landslide deposits and soil can improve the nutrient content of purple soil.