PURPOSE: To reduce ethanol reflux from the needle channel by injecting rat blood immediately after theinjection of ethanol into rat liver. MATERIALS AND METHODS: The first experiment involved 33 rat livers whichwere divided into four groups (three livers in group 1 ; ten in groups 2, 3, and 4). Group 1 animals were used ascontrols, and 0.1ml saline was injected into the liver ; in group 2, ethanol-Tc-99m-O4- mixed solution (0.1 ml,0.2 mCi) was injected into the liver ; in groups 3 and 4, the needle channel was blocked with 0.02 ml of freshblood and old blood, respectively, after the injection of ethanol. After removing the needle, a 3cm round filterpaper was laid on each injection site to absorb refluxed ethanol-Tc-99m-O4- mixed solution from the liver, andeach paper was then counted by a gamma camera unit. In the second experiment, 33 rats were divided into fourgroups (three rats in group 1 ; ten in groups 2, 3, and 4). Group 1 animals were used as controls, and afterexposing the left lateral lobe of the liver, 0.05 ml of saline was injected; in group 2, 0.05 ml of ethanol wasinjected into the liver in groups 3 and 4 the needle channel was blocked with 0.02 ml of fresh blood and oldblood, respectively, after the injection of ethanol. After ten days, peritoneal adhesions were scoredmacroscopically and microscopically. RESULTS: In the first experiment using ethanol-Tc-99m-O4- mixed solution,groups blocked with blood after the injection of mixed solution showed lower gamma counts than the group injectedwith mixed solution only (p-value = 0.0002). The group blocked with old blood showed the lowest count.Macroscopical and microscopical examination of peritoneal adhesions indicated that the grade of adhesion was lowerin groups blocked with blood than in the group injected with ethanol only (p-value = 0.0261 and 0.0163,respectively). CONCLUSION: The above results suggest that an injection of blood after an injection of ethanol isa very effective way of preventing reflux from the liver.