To investigate the effect of sand content on the growth of slope ecological restoration plants, we used the Analytic Hierarchy Process to select suitable plant species and then conducted planting experiment with sand added in Yili loess slopes. By analyzing changes in plant coverage, cumulative soil evaporation, and maximum crack rate, we found that adding sand to the planting soil can speed up plant germination and improve the soil germination rate. During early growth stage with sufficient water supply, the germination rate and plant coverage are positively correlated with sand content. Nevertheless, under drought conditions, plant coverage decreases as sand content rises. The cumulative evaporation of soil moisture is positively correlated with sand content and varies significantly with temperature fluctuations. Higher temperatures lead to larger differences in cumulative evaporation among samples with different sand contents, but these differences gradually narrow as the temperature drops. Taking 40% sand content as the threshold for optimal conditions: when sand content is below 40%, it is positively correlated with the maximum crack rate, and an increase in the maximum crack rate corresponds to an increase in the peak plant coverage of each sample. However, when sand content exceeds 40%, sand content and the maximum crack rate display a negative correlation. As maximum crack rate increases, the peak plant coverage of sample decreases. For wild or poorly maintained ecological restoration sites, an optimal sand content of 20% is recommended. For artificially maintained ecological restoration sites, a 60% sand content is optimal. In flat, human-intervened ecological restoration and maintenance sites, full sand coverage is the best choice.