Objectives 1) To explore the pattern of CD147 expression and its applicability in the prognosis of liver hepatocellular carcinoma (LIHC). 2) To establish hazard ratio and probability points for predicting the prognosis of HCC by correlating CD147 expression with clinical characteristics. 3) To determine if CD147 can be a reliable biomarker in LIHC prognosis. Methods CD147 expression profile in LIHC and corresponding clinical data were obtained from The Cancer Genome Atlas (TCGA) database. The expression patterns of CD147 were then validated by analyzing data from the Gene Expression Omnibus (GEO) database. In addition, CD147 Immunohistochemistry (IHC) in LIHC was obtained from the Human Protein Atlas (HPA). CD147 expression patterns and clinical characteristics in the prognosis of HCC were analyzed by accessing the UALCAN web resource. Accuracy, sensitivity, and specificity of the CD147 expression profile in predictive prognosis were determined by the Time-dependent Receiver Operating Characteristic (ROC) curves. Kaplan-Meier curves were plotted to estimate the hazard ratio (HR) of survival in LIHC. Univariate and multivariate Cox regression proportional hazards analyses of CD147 expression levels and clinical characteristics as prognostic factors of LIHC were performed. Nomograms were used to establish probability points and predict prognosis. Results Data from TCGA and GEO databases revealed that CD147 was significantly overexpressed in LIHC (p = 1.624e-12 and p = 1.2e-05), respectively. The expression of CD147 and prognosis of LIHC were significantly correlated with the clinical characteristicsof LIHC as per the data from the UALCAN web resource (p < 0.05). Kaplan-Meier analysis of CD147 expression in LIHC revealed that the high expression groups showed poor prognosis and an HR of survival > 1 [(Log-rank test, p = 0.000542, HR (in high expression groups) = 1.856, 95% CI (1.308, 2.636)]. ROC curves were plotted to analyze the 1-year, 3-year, and 5-year survival rates, the Area under the ROC curve (AUC) values were 0.675, 95% CI (0.611–0.74); 0.623, 95% CI (0.555–0.692); and 0.664, 95% CI (0.582–9.745) respectively. Univariate Cox analysis of CD147 expression and clinical characteristics of LIHC, and multivariate Cox analysis of CD147 patterns and pTNM-stage showed significant differences [(Uni-Cox, p = 0.00013, HR = 1.42437, 95% CI (1.8838, 1.70723) and p = 0.00066, HR = 1.37612, 95% CI (1.14521, 1.65359); Multi-Cox, p = 0.00578, HR = 1.50746, 95% CI (1.12637, 2.0175) and p= 0.00336, HR = 1.44319, 95% CI (1.12941, 1.84415)]. Nomograms were plotted to establish the probability points and predict prognosis, the total points ranged from 0 to 180, and the C-index value was 0.673, 95% CI (0.6–1.0), p < 0.01. Conclusions Overexpression of CD147 was correlated with poor prognosis in LHIC. CD147 expression profile combined with clinical characteristics can reliably predict the prognosis of HCC. CD147 can serve as a biomarker to predict the prognosis of HCC.