Abstract: To better utilize a recombinant inbred line（RIL）population derived from the cross between Shuhui527 and Mianhui436 in breeding programs, we established genomic selection（GS）models for this population and identified the optimal model for future screening of superior restorer lines to enhance breeding efficiency. Following comprehensive phenotyping of agronomic traits and genotyping of the population, three GS models were developed. The predictive accuracy and performance of these models across multiple traits in the RIL population were systematically evaluated to explore their applicability. The results showed that: eight quantitative traits in rice - heading date, productive tillers per plant, plant height, grain length, grain width, panicle length, grain number per panicle, and yield per plant - exhibited polygenic inheritance patterns, making them suitable for GS analysis. High-throughput sequencing generated genotype data yielding 454,129 filtered SNPs from 181 accessions, which were used for model construction. Comprehensive evaluation based on model accuracy and predictive capacity revealed that the GBLUP model exhibited superior overall performance, demonstrating its potential for practical implementation in subsequent breeding applications.