Recently, RNA has shown great potential for use in forensic genetics. Our previous work indicated that hair shafts possess detectable RNA levels. Fallen hair samples at crime scenes are common, but human identification is difficult because of the degeneration of traceable nuclear DNA. We aimed to establish a new mRNA polymorphism assay for the identification of hair shafts in humans. In this study, we utilized polymorphic mRNAs to obtain a human identification profile that is more detailed than that of any previously reported method. Ten to fifteen pieces of 5-cm hair shafts were used to extract total RNA from 40 individuals. RNA was transcribed into cDNA and typed on a BGISEQ T7 platform using a massively parallel sequencing assay encompassing 404 coding region and untranslated region (UTR) single nucleotide polymorphisms (SNPs) from 78 genes. The multiplex assay was evaluated for sensitivity, species specificity, capability for aged hair shafts, consistency of the typing results for hair borne from different body parts, and genomic DNA (gDNA)/mRNA of the same individual. We also obtained the genetic parameters for human identification in a Chinese population. Genes that did not meet this threshold were excluded from the analysis. Ultimately, 71 genes containing 284 SNPs, amplified with 228 amplicons were retained. Polymorphisms were observed in 210 amplicons. The random match probability (RMP) values ranged from 1.24 × 10-44 to 1.14 × 10-72 (median = 4.36 × 10-50). When one piece of 5-cm hair shaft was used, 70.31-72.05 % of the amplicons could be detected, and 46.72-62.01 % of the amplicons showed the same genotype as 15 pieces of hair shafts. A total of 44-82 amplicons were detected in hair shafts from four common animals (cats, dogs, rabbits, and rats). However, the genotyping of most SNP/microhaplotype (MH) markers was inconsistent with the database records. This study provides a new strategy for human identification of hair shafts.