This review was undertaken to examine the evidence of effectiveness of high-dose chemotherapy and autologous bone marrow transplantation (HDC/ABMT) for the treatment of metastatic breast cancer and to compare the magnitudes of the benefits and harms of HDC/ABMT with those of conventional doses of chemotherapy. Published studies were reviewed and analyzed. No randomized controlled trials have been published that evaluate HDC/ABMT. Only one internally controlled study has been conducted; it demonstrated that HDC/ABMT and conventional treatment have virtually identical outcomes. Comparisons of uncontrolled clinical series are confounded by patient selection and other biases. Gross comparisons indicate that, compared with conventional-dose chemotherapy, HDC/ABMT achieves (1) higher complete response rates (36% v 8%), (2) higher overall response rates (70% v 39%), (3) similar median response durations (8 months v 9.6 months), (4) similar median survival durations (16 months v 16.6 months), and (5) similar overall survival rates (eg, 43% 2-year survival v 39%). Observations of cases with longer-term disease-free survival are promising but not conclusive. High-dose chemotherapy with ABMT has a higher treatment-related mortality rate (5% to 15% v 1%), a high rate of nonmortal toxicity (approximately 30%), and a high rate of side effects (approaching 100%). Firm conclusions are not possible because of the lack of controlled studies and the presence of numerous biases. However, the existing evidence does not demonstrate that HDC/ABMT is superior to conventional-dose chemotherapy for the treatment of metastatic breast cancer. Randomized controlled trials are needed.