The prognosis for patients with primary breast cancer involving multiple axillary lymph nodes is poor. Only about 30% of patients remain disease-free at 5 years from diagnosis despite surgery, conventional-dose chemotherapy, and radiation therapy. In nonrandomized studies, the use of high-dose chemotherapy as consolidation therapy after standard-dose induction chemotherapy has resulted in an apparent improvement in disease-free survival rates to over 70%. These results have prompted the National Cancer Institute to sponsor large-scale, multicenter, randomized comparative trials of this strategy. This Intergroup Study (Cancer and Leukemia Group B 9082, Southwest Oncology Group 9114, and National Cancer Institute of Canada MA13) compares two treatment strategies in women with primary breast cancer involving 10 or more axillary lymph nodes. Arms A and B are identical in the use of four cycles of conventional therapy with cyclophosphamide and doxorubicin and fluorouracil, radiation therapy, and tamoxifen. The only difference between the two arms is the dose intensity of the cyclophosphamide, cisplatin, and carmustine given following conventional adjuvant treatment. Arm A dictates bone marrow, peripheral blood stem cell, and hematopoietic growth factor support and frequently requires a prolonged hospital stay with high resource utilization. Arm B, with its less dose-intensive therapy, requires considerably less support to apply the treatment. Because of the high cost of this therapy and the requirement for technology-intensive support, there has been considerable interest in economic outcome assessments.