Hepatocellular carcinoma (HCC) has emerged as a primary health problem and threat to global mortality, especially in China. Since pyroptosis as a new field for HCC prognosis is not well studied, it is important to open a specific prognostic model. In this study, consensus clustering method for 42 pyroptosis-related genes to classify 374 HCC patients in the TCGA database. After cox regression analysis of the differentially expressed genes between the two clusters, LASSO-Cox analysis was then performed to construct a pyroptosis-related prognostic model with 11 genes including MMP1, KPNA2, LPCAT1, NEIL3, CDCA8, SLC2A1, PSRC1, CBX2, HAVCR1, G6PD, MEX3A. The ICGC dataset was served as the validation cohort. Patients in the high-risk group had significantly lower overall survival (OS) rates than those in the low-risk group (p < 0.05). COX regression analysis showed that our model could be used as an independent prognostic factor to predict prognosis of patients and was significantly correlated with clinicopathological characteristics. Nomogram showing the stability of the model predicting the 1, 3, 5 year survival probability of patients. In addition, based on the risk model, ssGSEA analysis revealed significant differences in the level of immune cell infiltration and activation of immune-related functional pathways between high and low-risk groups, and patients with the high-risk score may benefit more from treatment with immune checkpoint inhibitors. Furthermore, patients in the high-risk group were more tend to develop chemoresistance. Overall, we identified a novel pyroptosis-related risk signature for prognosis prediction in HCC patients and revealed the overall immune response intensity of the tumor microenvironment. All these findings make the pyroptosis signature shed light upon a latent therapeutic strategy aimed at the treatment and prevention of cancers.