This experiment was designed to compare the effect of two selenium sources at the dosage of therapeutic level on hepatocarcinogenesis and angiogenic cytokines in DEN-induced hepatocarcinoma rats to further approach their possible anticancer's mechanism. One hundred and seventy-eight Sprague-Dawley (SD) rats (average weight being 100-120g) were randomly divided into 5 groups (I-V). Animals in group I, group II and group III served as the negative control, sodium selenite control (SS) and positive controls respectively, and received 0.1, 3.0, and 0.1mg/kg selenium from sodium selenite supplemented diets during the whole experimental time. Rats in group IV and group V were fed with selenium from selenium-enriched malt (SEM) and sodium selenite (SS) supplemented diets (3mg/kg respectively). To balance the nutritional content among each group, normal malt which was not treated with selenium was added into the diets of the challenge groups. The nutrition contents, except the selenium of the diet in each group, were similar and in accordance with NRC standards. Rats in groups III-V were treated by aqueous diethylnitrosamine solution (100mg/L) at the dosage of 10mg/kg body weight every day for 16 weeks to induce hepatocarcinoma, and drank sterilized water for an additional two weeks. Rats in group I and group II drank sterilized water throughout the experiment. At 4th, 8th, 12th, 16th week, five rats in each group were then sacrificed by cervical decapitation. At the termination of the study, at 18th week, the surplus rats were sacrificed by cervical decapitation. Feed was withheld from the rats for 12h before sampling. The number of hepatoma nodules in liver and mortality of rats were calculated. The values of the following items, including α-fetoprotein (AFP), gamma-glutamyltranspeptidase (GGT), tumor necrosis factor-α (TNF-α), insulin-like growth factors-II (IGF-II), nitric oxide (NO) and total nitric oxide synthase (T-NOS) in plasma were determined. At the same time, the positive numbers of vascular endothelial growth factor (VEGF) and protein kinase C-α (PKCα) staining cells in tumor tissue were analyzed by immunohistochemistry using the Envision two step methods with a kit. The results indicated that SEM could significantly decrease the mortality of rats and the number of hepatoma nodules, values of GGT and AFP, and the levels of IGF-II, NO and NOS and lessen the positive numbers of VEGF and PKCα staining cells in tumor tissue. Moreover, SEM could increase the levels of TNF-α in the initiated time of hepatocarcinogenesis, whereas, decrease the levels of TNF-α in the progressive time of hepatocarcinogenesis. SS could only significantly inhibit the positive numbers of PKCα staining cells in tumor tissue, decrease the levels of GGT, AFP and TNF-α at minority sampling times, and increase the levels of NO. In conclusion, SEM could reduce the mortality. It might be related to deaden significantly the lesion of liver, delay the cause of hepatocarcinogenesis, and inhibit the progress of angiogenesis to increase the livability of DEN-induced hepatocarcinoma rats. SS at the same therapeutic dosage had less effect on the hepatocarcinogenesis by inhibiting angiogenesis and relative cytokines to some extent.