This study used a large-scale cancer database in determining the survival prognostic factors among primary liver cancer (PLC) subjects. A total of 28,939 subjects diagnosed with PLC were analysed. Survival estimates were performed with Kaplan-Meier methods. Cox's proportional-hazards model estimated the death risk (hazard ratio (HR)) of prognostic factors. The prognostic indicators associated with higher risk of all-cause deaths are male gender (males versus females; HR=1.16, 95% confidence intervals (CI), 1.13-1.20), diagnosis at later period (shown in 1990-1994 versus 1985-1989; HR=1.04, 95% CI, 1.01-1.08), increasing age at diagnosis, subjects with adenocarcinoma/cholangiocarcinoma (CC) and with no therapy against those with chemotherapy. The overall 5-year survival rate for all causes of death was significantly poorer in males (13.7%) than females (17.2%). Subjects diagnosed with hepatoblastoma and treated by surgical resection alone had superior prognosis. Particularly, subjects with adenocarcinoma and CC were more likely to die in other metastatic cancer.