Background: The main reason for intravenous immunoglobulin G (IVIG) treatment in acute phase of immune thrombocytopenic purpura (ITP) patient is prevention of possible life-threatening hemorrhage with rapid increment of platelet by Imbach since 1981. All doctors followed him with a little modification until now. Methods: All acute ITP patients were admitted and checked daily platelet count. During 30 years we treat childhood ITP with IVIG according to individual response with 4 different methods. I compared the clinical response, the time to reach a target platelet count and the total dose with prospective study. Results: On the 1st trial all patient were well responded with 5 days treatment (Korean J Pediatr. 28:483-91, 1985). On the 2nd trial, the mean treatment days was 2.8 days to raise above 100,000/μL, but the platelet count was not increased above 100,000/μL in 11 with full dose. The relapse and the chronic ITP was less developed in rapid responders (RR: 1-3 treatment days) at the level of 50,000/μL than in slow responders (SR: more than 3 days) ( P0.05) (Korean J Hematol 2001;36:241-6). On the 3rd trial, the platelet counts began to rise above 50,000/mm3 at 2.9 days. Four of 29 RR (treatment days; 1-2), 5 of 24 intermediate responder (IR: 3-4) and 7 of 15 SR (5 or more) were relapsed (p value of RR vs SR=0.028). Seven of 28 RR, 1 of 11 IR and 6 of 8 SR were chronic type (p value of RR vs SR=0.016, IR vs SR=0.006) (Korean J Hematol 2001;36:247-52). On 4th trial, the mean age was 4 year and 7 months (2 mo-14.5 yr) and mean platelet count was 8,000 (2,000-30,000)/μL. Mean 2 days treatment were needed to increase the platelet count 50,000/μL. In comparison between non re-treat (18) and re-treat (8) group, the mean age, platelet count at diagnosis and at the end of IVIG, total doses and time to over 50,000/uL had no statistical difference. Although all 10 non re-treat group who were follow-up over 6 months were acute, all 3 re-treat group who were follow-up over 6 months were chronic (p0.003) (Clin Pediatr Hematol Oncol 2006;13:143-9). No one died with 4 studies. Conclusions: The IVIG dose was different according to one’s clinical response. In some patient only 1/10 of usual doses (2 g/kg) were sufficient to reach a safe platelet count. We propose a group study for clinical response according to IVIG dose and early prediction of chronic ITP.