The goal of marker-assisted backcrossing is to reduce the number of generations significantly by using genome-based molecular markers. Among other types of molecular markers, SNP (single nucleotide polymorphism) is mostly used in genetic diversity analysis due to its abundance. To develop high-throughput SNP marker for MAB system, we selected 20 Chinese cabbage lines each represent traits as inner leaf color, disease resistance, head type and maturity etc. Then, we sequenced the transcriptomes of 20 lines by using Illumina Hiseq2000. The average transcriptome size was 1.37 Gbase, and the average of short reads mapping rate was about 62.15% (30xcoverage). We identified 13,976 SSR markers and 380,198 SNPs by aligning contigs of 20 Chinese cabbage lines. To develop SNP marker set, we chose 409 SNPs that covers the whole Brassica rapa transcriptome. The filtering criteria were depth, polymorphism, segregation ratio, lack of adjacent SNP and copy number. We positioned the selected SNP markers to the Chinese cabbage linkage map. Clustering dendrogram was produced using SNP marker and three different clusters were generated. The result showed that the genotyping data is partially linked to the phenotyping data. We assume that the developed SNP marker set can be applied in the Chinese cabbage MAB system soon.