The primary goal of therapy in patients with metastatic breast cancer is palliation and prolongation of life with quality. Although there are no randomized trials comparing chemotherapy with supportive care in women with metastatic breast cancer, chemotherapy clearly provides a survival benefit in metastatic breast cancer. Due to diagnosis at earlier phases of metastatic disease and more effective therapies, the median survival of patients treated with modern taxane-based chemotherapy is now reaching approximately 2 years. The duration of chemotherapy in patients responding or stable disease remains controversial, since quality of life is not usually adversely affected and may even be improved in many patients receiving cytotoxic chemotherapy. In addition, many commonly used chemotherapeutic agents are not limited by cumulative toxicity in metastatic breast cancer patients. Several studies investigated the role of maintenance chemotherapy suggest that maintenance chemotherapy is associated with superior time to progression but no survival gain. However, these randomized trials did not incorporate taxane-based chemotherapeutic regimens, the new standard of care in metastatic breast cancer patients these days. A 1998 metaanalysis of 1,986 patients randomly assigned between combination chemotherapy and single-agent therapy in metastatic breast cancer patients demonstrated a survival advantage to combination chemotherapy, with a hazard ratio of 0.82 (range, 0.75 to 0.90). Although there are several randomized trials showing negative results for survival gain in patients who received maintenance chemotherapy, the role of maintenance chemotherapy with newer agents, such as docetaxel or paclitaxel, have not been established yet. Although Italian MANTA trial demonstrated no difference in PFS or survival between the two arms, their metaanalysis advocated survival benefit of maintenance therapy. Since gemcitabine/paclitaxel (GP) combination chemotherapy is one of the two regimens which showed definite survival gain with favorable toxicity from a randomized trial, we plan to randomize patients who responded to six cycles of GP induction chemotherapy to receive additional maintenance GP chemotherapy until disease progression versus observation. We hypothesize that patients who receive maintenance GP chemotherapy will do better in terms of progression free survival.