1097 Background: Docetaxel and vinorelbine as single agents have both demonstrated efficacy in the treatment of metastatic breast cancer. This prospective study compared the efficacy of these two treatments in anthracycline refractory metastatic breast cancer, and assessed the disease response when used in cross over after disease progression on the initial regime. Methods: Patients with metastatic breast cancer who had progressed following anthracycline treatment were randomly assigned to either docetaxel 100mg/m2 on day 1 every 21 days or vinorelbine 25mg/m2 day 1 every 14 days. Patients were eligible to switch over to the other regime at progression. Results: 38 patients were randomized, 18 to docetaxel and 20 to vinorelbine, 4 patients became ineligible as a result of protocol violations or withdraw from study (1 in docetaxel, 3 in vinorelbine), and response data was not available on 1 patient randomized into docetaxel arms. The objective response was 12.5% (PR) in the docetaxel arm with 31% obtaining clinical benefit from stable disease (SD). In the vinorelbine arm there was a 6% PR with 6% SD. The median time to progression was 10.8 weeks (range 3-39 weeks) in the docetaxel arm and 9.5 weeks (range 4-37 weeks) in the vinorelbine arm. At progression 37.5% (6 patients) switched from docetaxel (T) to vinorelbine (V) and 53% (9 patients) from vinorelbine (V) to docetaxel (T). There was 1 objective response in V→T but none in T→V. The median time to progression was 9.5 weeks (2-16 weeks) in T→V and 15 weeks (range 7-43 weeks) in V→T. Grade 3 or 4 haematological toxicities were more common in the docetaxel arm (11) compared to the vinorelbine arm (1). Conclusions: Docetaxel and vinorelbine have similar efficacy as single agents in anthracycline refractory disease, however vinorelbine is better tolerated. Whilst an objective response was seen only in the crossover of V→T, clinical benefit was observed in both V→ T and T→V crossover. No significant financial relationships to disclose.