To provide a scientific reference for the prevention and treatment of pyrazinamide-resistant tuberculosis (PZA-R TB), we analyzed the prevalence and risk factors of pyrazinamide-resistant tuberculosis in Hunan province and described the genotyping and clustering characteristics of the pyrazinamide-resistant (PZA-R MTB) isolates. The drug susceptibility test results of first-line anti-tuberculosis drugs including isoniazid (INH), rifampicin (RFP), streptomycin (SM), ethambutol (EMB) and pyrazinamide (PZA), and the characteristics of patients were collected from 3 862 tuberculosis patients in Hunan Chest Hospital (Institute of Tuberculosis Control and Prevention) from January 2016 to December 2018. The prevalence of PZA-R TB was calculated and risk factors were analyzed by univariate and multivariable logistic regression analysis. Two hundred and twelve isolates selected from June 2017 to June 2018 were genotyped using the 24-loci MIRU-VNTR system. The genetic difference value (h), and the Hunter-Gaston index (HGI) were used to evaluate the resolution and variation for the 24 loci. MIRU-VNTR results were analyzed using BioNumerics 5.0 software to conduct cluster analysis. Clustered isolates were further analyzed by gene sequencing. The rate of PZA-R TB among tuberculosis patients and MDR patients was 14.7%(566/3 862) and 60.5%(511/844), respectively. Multivariable logistic regression analysis showed that patients who were INH mono-resistance and MDR had a higher risk of developing PZA resistance, compared with TB patients who were pan-sensitive to anti-TB drugs (INH, RFP, SM, and EMB). The adjusted value (95%) was 13.08(5.67-30.18), 298.41(164.88-540.08), respectively, and values were all less than 0.01. Clustering analysis showed that 65 strains formed 19 clusters, the clustering rate was 30.7%(65/212). Of 19 clusters, eight clusters had at least two isolates with identical mutation types within each cluster. In eight clusters, cluster 4, 6, 16 had four, three, and two patients who lived in the same county, respectively, thus providing probable epidemiological links for the recent transmission of PZA-R . At least 47.6%(101/212) of PZA drug-resistant TB patients were suggestive of primary drug resistance caused by transmission. The prevalence of PZA-R TB was severe in Hunan province. PZA susceptibility testing should be performed for isolates resistant to any first-line anti-tuberculosis drugs, especially for MDR-MTB isolates. Nearly half of tuberculosis patients were suggestive of primary drug resistance caused by transmission. The prevention and treatment strategy of PZA-R TB should focus on the standardized treatment and management of patients as well as control of the source of infection.