This retrospective analysis of 254 children less than 15 years of age treated with MCP-841 protocol from June 1992 to June 2002 was undertaken to identify the pattern of relapse and determine management lacunae. Two hundred twenty-three (87.8%) children achieved a complete remission of whom 40 (17.9%) relapsed. The mean age of relapsed patients was 6.5 years. The male/female ratio was 9:1. There were 23 (57.5%) isolated bone marrow (BM), 7 (17.5%) isolated central nervous system (CNS), 2 (5%) isolated testicular, 5 (12.5%) BM+testes and 1 each of BM+CNS, CNS+testes, and isolated bone relapses. Twenty-seven children (67.5%) relapsed on-therapy whereas 13 (32.5%) relapsed posttherapy. All 9 CNS relapses occurred on-therapy whereas 5/8 (62.5%) of testicular relapses occurred posttherapy. Lymphadenopathy was the only significant predictor for relapse. High-risk features such as age less than 1 year and greater than 10 years (P=0.047) and white cell count greater than 50.0 x 10(9)/L (P=0.044) were significantly more frequent in patients with early on-therapy relapse than in patients with off-therapy relapse. The overall survival in the entire study cohort was 67+/-3.5%. Modest survival outcome, relapse while on chemotherapy and the higher incidence of CNS and testicular relapse indicate the need for reappraisal of our treatment protocol. There is a need of identifying risk factors and high-risk groups in our set of patients and risk-stratified intensification of chemotherapy in them.