Intractable pain from advanced carcinoma of the upper abdomen is difficult to manage. On method used to control pain associated with these malignancies is to block off the splanchnic nerve. In 1919 kappis described a technique by which the splanchnic nerve of the upper abdomen could be anesthetized, using a percutaneous injection. This method has been used for the relief of upper abdominal pain due to hepatoma and cancer of the pancreas, stomach, gall bladder, bile duct, and colon. During the period from November 1968 to January 1986, this method was used in 208 cases of malignancy at Severance Hospital and clinically evaluated. Patients were retroactively grouped according to the stage of development of technique used. Twelve patients who received the treatment in the period from November 1968 to March 1977 were designated as group 1, 26 patients from April 1977 to April 1979 as group 2, and 170 from May 1979 to January 1986 as group 3. The results are as follows: 1) The number of patients receiving splanchnic nerve block has been increasing since 1977. 2) A total of 208 patients, including 133 males and 75 females, ranging in age from 18 to 84 and averaging 51. 3) The causes of pain were stomach cancer 90, pancreatic cancer 69, and miscelianeous cancer 49 cases respectively. 4) There were 57.7% who had had surgery, and 3.7% of whom had had chemotherapy before the splanchnic nerve block was done. 5) These blocks were carried out with the patient in the prone position as described by Dr. Moore. Fro group 2 and 3 C-arm image intensifier was used. In group 1, a 22 gauze 10cm long needle was inserted at the lower border of the 12th rib on each side about 7cm from the midline. The average distance from the middle was 6.60±0.61cm on the left side and 6.60±0.83cm on the right side in group 2, and 5.46±0.76 on the left side and 5.49±0.69cm on the right side in group 3. The average depth to which the needle was inserted was 8.60±0.52cm on the left side and 8.74±0.60cm on the right side in group 2, and 8.96±0.63cm on the left side and 9.18±0.57cm on the right side in group 3. 6) The points of the inserted needles were positioned in the upper quarter anteriorly. 51.8% on the left side and 54.4% on the right side of the L1 vertebra by lateral roentgenogram in group 3. The inserted needle points were located in the upper and anterolateral part, of the L1 vertebra 68.5% on the left side and 60.6% on the right side, on the anteroposterior rentgenogram in group 3. The needle tip was not advanced beyond the anterior margin of the vertebral body. 7) In some case of group 3, contrast media was injected before the block was done. It shows the spread upward along the anterior margin of the vertebral body. 8) The concentration and the average amount of drug used in each group was as follows: In group 1, 39.17±6.69ml of 0.5% -1% lidocaine or 0.25% bupivacaine were injected for the test block and one to three days after the test block 40.00±4.26ml of 50% alcohol was injected for the semipermanent block. In group 2, 13.75±4.48ml of 1% lidocaine were used as the test block and followed by 46.17±4.37ml of 50% alcohol was injected as the semipermanent block. In group 3, 15.63±1.19ml of 1% lidocaine for test block followed by 15.62±1.20ml of pure alcohol and 16.05±2.58ml of 50% alcohol for semipermanent block were injected. 9) The result of the test block was satisfactory in all cases. However the semipermanet block was 83.3 percent of the patients in group 1 who received relief from pain for at least 2 weeks after the block, 73.1% in group 2, and 91.8% in group 3. In these unsuccessful cases, 2 cases in group 1 were controlled by narotics but 7 cases in group 2 and 14 cases in group 3 received the same splanchnic nerve block 1 or 2 times again within 2 weeks. But in some cases it was 3 to 5 months before the 2nd block and in 1 cases even 7 years. 10) The most common complications of splanchnic nerve block were hypotensino(25.5%) occasional flushing of the face, nausea, vomiting, and chest discomfort. 11) For the patients in group 3, the supplemental block most commonly used was a continuous epidural block; it was used as a diagnostic block and to afford relief from pain before the splanchnic nervel block was done. 12) The interval between the receiving of the alcohol block and discharge was from 5 to 8 days in 61 cases(31.1%) and from 1 to 2 days in 48 cases(24.5%). From the above results, it can be concluded that the splanchnic nerver block done in the prone position with pure and 50% alcohol immediately after an effective test block with 1% lidocaine under C-arm fluoroscopic control is satisfactory and reliable. How to minimize the repeat block is still a problem to be solved.