Laparoscopic splenectomy has been applied in an effort to decrease the perioperative morb idity and mortality, postoperative discomfort and length of hospital stay. At its current primitive stage of development, laparoscopic splenectomy is limited to the elective removal of normal sized spleen, and the complex vasculature and the many peritoneal attachments of the spleen seems to defy the laparoscopic approach. We have selected patients with idiopathic thrombocytopenic purpura (ITP) because the spleen is usually normal or small in size. Fifteen cases of laparoscopic splenectomy have been performed for ITP since October 1994. All the patients were female with age ranging from 23 to 47 years. We used 5 ports (three, 10 mm ports and two, 5 mm ports) for the first 8 cases, but we found we could eliminate one 10 mm port after changing the patient's position from supine to the right lateral kidney position. There were no cases of conversion to explolaparotomy in this series. The mean operation time was 3 hours, 20 minutes. Vascular control was achieved using endovascular cli ps and an average of 29 clips were used per case. For the removal of the dissected spleen, a sterile vinyl bag was introduced into the peritoneal cavity and the spleen was placed in the bag. Piecemeal removal of splenic tissue from the bag was accomplished through the umbilical port. Mean hospital stay was 5.2 days. There was no significant perioperative morbidity or mortality associated with the surgery. Among the 15 patients who underwent laparoscopic splenectomy, 12 patients had been off steroids with platelet counts greater than 120,000 /ul, 2 patients had been on small dose of steroids with platelet counts between 50,000 and 120,000 /ul, and one patient have been on dame dose of steroid with no response. Conclusion: Laparoscopic splenectomy is a relative safe and reasonable operative procedure for patients with ITP.