The aim of this study is to investigate the value of pre-treatment 18F-FDG PET/CT in predicting the pathological characteristic of HCC and recurrence after liver transplantation (LT). A total of 34 patients who underwent 18F-FDG PET/CT before LT for HCC and did not receive any other treatment were retrospectively enrolled in the study. The maximal standard uptake value of the tumor (T-SUVmax), normal liver tissues (L-SUVmax), and mediastinal blood pool (B-SUVmax) were derived, followed by the calculations of the T-SUVmax/L-SUVmax (T/L) and the T-SUVmax/B-SUVmax (T/B) ratios. Combined with the post-transplantation pathological results and ROC curve, the performance in predicting the pathological characteristic and the recurrence after LT were evaluated. The AUCs for T-SUVmax, T/L, and T/B in predicting the pathological grade of tumors were 0.820, 0.784, and 0.806, respectively. Alternatively, the AUCs for T-SUVmax, T/L, and T/B in predicting the recurrence after LT were 0.865, 0.899, and 0.931, respectively. The individual cutoff values were 5.0, 1.7, and 2.2, corresponding to a predication accuracy of 88.2%, 85.3%, and 88.2%, respectively. In addition, the AUCs for T/B in predicting microvascular invasion (mVI) and liver capsular invasion (LCI) were 0.825 and 0.788, respectively. The pre-treatment 18F-FDG PET/CT is effective for predicting recurrence of HCC after LT. In addition, it demonstrates values for predicting the pathological characteristic of HCC such as pathological grade, mVI, and LCI.