The Cancer Therapy Evaluation Program, National Cancer Institute (CTEP, NCI) strongly supports the role of controlled clinical trials in improving the care of children with cancer, and particularly the central role that the pediatric Cooperative Groups play in this process. Trends that threaten the ability to perform these trials include the increasingly limited financial resources available for clinical investigations and the sentiment within some circles that controlled clinical trials may be inappropriate for ethical reasons. The inherent risks of accepting a new therapy without rigorous comparison to existing therapy strongly support the need for randomized trials with adequate accrual to answer important therapeutic questions in a timely and reliable fashion. Retrospective analysis of multiple clinical trials is one method for identifying compelling hypotheses to be tested prospectively. Using this method, we have demonstrated the association between doxorubicin dose intensity and positive response and outcome for patients with Ewing sarcoma and osteosarcoma, thereby providing direction for the selection of important therapeutic questions to be addressed in future clinical trials for these malignancies.