To clarify the infection source and genetic characteristics of Giardia lamblia isolated from a diarrheal patient, fresh fecal sample of the patient was collected. Microscopy, real-time fluorescent quantitative PCR and nested PCR were comprehensively applied for pathogen detection. Metagenomic sequencing and analysis were performed targeting the small subunit ribosomal RNA (SSU rRNA), triosephosphate isomerase (tpi), glutamate dehydrogenase (gdh), β-giardin (bg) and translation elongation factor 1α (ef1α). Multilocus genotyping and genetic evolution analysis were carried out via sequence alignment and phylogenetic tree construction. Both microscopy and real-time fluorescent quantitative PCR confirmed Giardia infection. Only the gdh and ef1α genes were successfully amplified by nested PCR, and the SSU rRNA gene sequence was successfully detected by metagenomic sequencing. Multilocus genotyping showed that the isolate DXH001 obtained in this study belonged to assemblage B, and no mixed infection was detected. Phylogenetic analysis indicated that the gdh gene of isolate DXH001 shared 100% sequence identity with the human-derived strain EU847734 from Belgium, belonging to subtype BⅣ; the ef1α gene shared 98.43% sequence identity with the human-derived isolate HQ179598 from Australia, and the SSU rRNA gene shared 90.45% sequence identity with the human-derived strain U09491 from the United States, both clustered into the assemblage B branch. Compared with the human-derived reference sequence U09491, the SSU rRNA gene had 35 base substitutions and 5 deletions, and the sequence difference was mainly related to the insufficient sequencing data volume caused by the low abundance of the target sequence in the sample; the variations of gdh and ef1α genes were relatively conservative, which did not cause significant functional changes. In this study, the isolate DXH001 was identified as G. lamblia assemblage BⅣ via multilocus genetic analysis, suggesting the trans-regional transmission risk of this subtype. Metagenomic sequencing shows high sensitivity in low-abundance target detection, which can serve as an effective supplement for conventional molecular genotyping.