Eventhough the success ratio of the splenectomyis equal to that of steroid therapy (1), and splenectomy (2), including that by the laparoscopic method (3, 4) has recently been reported to be effective, splenectomy is not generally selected as the first choice of therapy for immunethrombocytopenic purpura. On October 16, 1995, we performed a laparoscopic splenectomy in preference to steroid therapy on a 15-year-old male field athlete with immunethrombocytopenic purpura. His platelet count on March 29, 2001 was 217,000/mm3 (217x107/) which confirmed our selection of laparoscopic splenectomy instead of the more commonsteroid therapy. The patient presented at our hospital with thrombocytopenia and nasal bleeding. Incidentally, platelet counts 3 months before his admission were 23,000-30,000/mm3 (23-30x107/). Nasal bleeding occurred occasionally, such as after afternoon exercise on the track or after taking a bath, and was difficult to stop. Other than positive platelet associated IgG and low platelet counts, the results of the patient's hematological tests were normal. The chronic nature of his thrombocytopenia was reflected in the commentof a local practitioner who noted that 1 year before the boy's visit to our hospital, his platelet count was"only 60,000/mm3 (60x107/)." In addition to being a member of the track and field club of his school, he was a record holder of the 200-meter sprint in his prefecture (Saitama Prefecture with population of 6 million). Because of his athletic aspirations, he and his parents were reluctant for him to undergo steroid therapy. Even though major disadvantages of splenectomy include rare but serious postoperative infections (5), the financial burden, and perioperative risks, splenectomy seemed similar to steroid therapy in that both therapies are palliative rather than curative. Complete remission of the disease, defined by platelet count, mean platelet life, and platelet production, is reportedly obtained only after splenectomy and not after prednisone treatment (6). The laparoscopic approach was selected because this induces less damageto abdominal muscle than the laparotomy approach. After the patient's parents gave informed consent, laparoscopic splenectomy following administration of high doses of immunoglobulin was performed under general anesthesia. The operation lasted for 7 hours and 30 minutes. Although this was significantly longer than expected, the spleen was completely resected without any dissemination within the abdominal cavity that might lead to accessory spleen formation. However, bleeding from splenic artery could not be stopped endoscopically, leading to partial abdominal fenestration. An unscheduled blood transfusion of 1,000 ml was necessary and, the admission period was 15 days 5 days longer than had been anticipated. The hospital fee (the patient's family is responsible for 30%and the health insurance system covers 70%) amounted to 540, 1 80 yen (equivalent to approximately US$5,000). At the beginning of the year, he resumed warm-up practices for sprinting ...