To investigate the effectiveness of thromboxane (Tx) synthase inhibitor in the prevention of experimental hepatic metastasis, an in vivo study was designed. Hepatic metastasis was brought about by injection of 1 x 10(5) cells of colon 38 tumor into the portal vein of C57 B1/65 mice. Seven groups (n = 16 in each group) received different treatments: with TxA2 synthase inhibitor (sodium ozagrel), 5, 10 or 15 mg/kg BW before tumor inoculation, and daily for the following 3 days, (groups A, B and C, respectively); with acetyl salicylic acid (aspirin), 1.0, 1.5 or 2.0 mg/kg BW (groups C, D, and E, respectively); a control group, inoculated with vehicle only. Serum TxB2, a stable metabolite of TxA2, and prostaglandin F1 alpha were measured. Labeling index for tumor proliferation by bromodeoxy-uridine radioimmuno-assay was also studied. Incidence of metastasis in groups A (60.5%), B (49.5%), C (43.0%), D (80.5%), E (66.0%) and F (58.4%) was less than that in the control group (100%). Tumor size, number of labeling index did not differ among the groups. Serum TxB2 (pg/ml) levels were significantly lower in all of the groups than in the control. Serum PGF1 alpha levels in the groups with aspirin were lower than those in sodium ozagrel. Tx synthase inhibitor is effective in the prevention of experimental hepatic metastasis when it is given before and immediately after tumor inoculation. As Tx synthase inhibitor leaves metabolic pathway to PGI2 production intact, it is more effective in the prevention of metastasis than aspirin since aspirin inhibits both thromboxane and PGI2.