<p>The protein content in soybean seeds serves as a crucial measure of soybean quality. Breeding high-protein varieties remains the most cost-effective and efficient approach to increasing soybean protein levels. Nevertheless, limited research has focused on identifying the genes responsible for high protein content among the diverse soybean cultivars. To address this gap, a genome-wide association study (GWAS) was conducted on 455 soybean varieties with varying protein content to predict and validate novel genes involved in regulating protein levels in soybean seeds. Protein content data were obtained from three distinct environments, along with three environmental variables derived from oil content, which is closely related to protein levels. Genotyping was performed using the SoySNP180k BeadChip, yielding genotype data for 63,306 non-redundant single nucleotide polymorphisms (SNPs). Five multi-locus GWAS methods were employed, resulting in the identification of 81 significant quantitative trait nucleotides (QTNs), of which 37 QTNs detected across different methods and environments were further analyzed. Moreover, the simulation platform Blib was used to conduct single-crossing simulation breeding on 81 QTN loci for actual breeding prediction. Haplotype analysis based on re-sequencing data confirmed 2 genes closely linked to protein synthesis, providing a theoretical basis for breeding high-protein soybean varieties and developing molecular breeding strategies.</p>