Introduction Around 5 to 20% of the population in the provinces of southern China are carriers of β-thalassemia mutations. In our hospital, more than half of couples treated with preimplantation genetic testing for monogenic diseases (PGT-M) are β-thalassemia carriers. The most common mutations are deletion in β-thalassemia genes (HBA1 and HBA2), including –SEA, -α 3.7, and -α 4.2. Preimplantation genetic haplotyping (PGH) is a powerful technology for PGT-M that can detect allele drop-out (ADO) accurately by detecting numerous linked or unlinked polymorphic markers in DNA segments surrounding the genes of interest. However, there is a lack of normative data references in the selection of linkage sites for disease. Therefore, a mutation-specific haplotype analysis will improve the efficiency of PGT-M for β-thalassemia. Materials & methods A total of 113 couples with β-thalassemia deletion mutations were collected from the Reproductive Medicine Center in the First Affiliated Hospital of Sun Yat-sen University and the Reproductive and Genetic Hospital of CITIC-Xiangya. A pre-examination validation for PGT-M was performed on these. The SNPs with large difference between reference frequency and alternate frequency in 452 haplotypes with or without three kinds of deletions were filtered out. Candidate sites close to the β-thalassemia genes were selected, upstream and downstream of genes within 1 Mb. Then, the haplotypes model was established with several specific genotypes of SNPs inherited with the mutations. Subsequently, the model was verified to predict the genotypes in another 34 couples after PGT-M. In addition, the ADO ratio of each site in 255 embryos was analyzed. Results Eight candidate SNPs (rs2562164, rs2857997, rs2857998, rs116995933, rs149153770, rs3918352, rs1203977, and rs1203979) were identified for the haplotypes model constructed for different genotypes. SEA-I (GCGTAACA), α3.7-I (AGACGACA) and α4.2-I (AGACGGAT) haplotypes were detected in 92.2%, 50.9%, and 70.8% genotypes with corresponding mutations, respectively. On average, 42.7% sites were able to be used as key SNPs in the verification of another 34 PGT-M cohort haplotypes and genotypes. The SEA-I model was detected in 91.1% of genotypes with –SEA mutation. The α3.7-I model was concordant with 37.5% of genotypes with the -α3.7 mutation. The α4.2-I models were inherited with 87.5% of genotype with the -α4.2 mutation. The average ADO ratio of the eight SNPs in the 255 embryos was 2.3%. Conclusions The haplotype analysis for common β-thalassemia deletion mutations including the eight-site model improves the efficiency of PGT-M in the population of southern China. The establishment of standardized diagnostic procedures and systems can be further extended to other populations for PGT-M in the future.