Abstract: In recent years, single nucleotide polymorphisms (SNP) have emerged as the preferred molecular markers for elucidating the genetic basis of crop-related traits. This preference is attributed to their advantages, including dense distribution across genomes, low mutation rates, and compatibility with automated sequencing technologies, which have significantly advanced genomic sequencing and genotyping capabilities. Identifying phenotypic formation genes independently of a reference genome can accelerate soybean improvement breeding. Traditional breeding methods predominantly rely on phenotypic selection, which involves screening based on observable external characteristics such as plant morphology, yield, and resistance. However, this approach is characterized by a long cycle, low efficiency, and susceptibility to environmental interference. In contrast, molecular marker-assisted selection (MAS) leverages the close linkage between molecular markers and genes that determine target traits. By detecting these markers, the presence of target genes can be inferred, thereby achieving the selection of desired traits and markedly enhancing breeding efficiency. The advent of high-throughput sequencing technology has facilitated the development of various efficient gene localization strategies. For instance, methods based on bulked segregant analysis (BSA), such as QTL-seq, Gradient-Seq, QTG-Seq, exome QTL-seq, and RapMap, compare genomic differences in populations with extreme phenotypes to rapidly identify candidate genes. Mutant-based approaches, including MutMap, NIKS algorithm, MutRenSeq, and MutChromSeq, integrate mutagenesis populations with sequencing technology to pinpoint functional genes. Additionally, techniques utilizing target sequence enrichment, such as RenSeq, AgRenSeq, and TACCA, enable efficient screening of SNPs associated with disease resistance. Furthermore, advancements in statistical methodologies, such as mixed linear models (MLM), multi-site genome-wide association studies (GWAS), and machine learning algorithms, have substantially improved the accuracy and efficiency of classical localization strategies and genome-wide association analyses. These innovations allow for more precise identification of micro-effect polygenic variations and rare alleles linked to complex traits, such as yield and stress resistance. To comprehensively review the progress of SNP marker applications in soybean breeding, this paper summarizes research achievements in 2024 regarding QTL localization and candidate gene function analysis for traits such as yield, quality, stress resistance, and plant architecture, aiming to facilitate the application of more effective SNP loci in molecular marker-assisted breeding.