PURPOSE: To determine the effects of blood vessels on the size and shape of microwave coagulation. MATERIALS AND METHODS: Microwave coagulation was performed with 60 W output and 60 second duration. In the first experiment five exvivo porcine livers were used to determine the size of the coagulation area and its reproducibility. The second experiment involved the used of two in-vivo porcine livers to determine how adjacent vessels affect the size and shape of coagulation. RESULTS: The result of the first experiment was that the maximum mean diameter of lesions was 1.4 cm +/- 0.1 , reproducible in the range of 1.3 c m -1 .5 cm. In the second experiment, maximum mean diameter was found to be 1.5cm +/- 0.1, reproducible in the range of 1.3 cm - 1.7cm, and the size and shape of the lesion was affected by nearby blood vessels. The shape factor of the lesion, defined as roundness of sphere, was 0.8, but the r a n g e ( 0 . 5 8 -0.92) was wide due to the effect of vascular cooling. This was more prominent in the portal vein than in the hepatic vein, and the minimum diameter of the portal vein which deformed the lesion by more than 1 mm was 0.1 mm. CONCLUSION: Microwave coagulation gives a well-defined lesion, the size of which can be reproduced, but size variation and nonuniformity can be caused by nearby blood vessels.