OBJECTIVE: Prognostic factors in 92 patients with brain metastases, treated in our hospital, are identified in order to determine subgroups of patients suitable for selection in future trials. METHODS: From January 1985 through March 1998, 117 patients with computed tomography(CT) or magnetic resonance(MR) image diagnosed brain metastases were referred to our department. The follow-up data of 92 cases were available in this study. One half of the patients were treated with conservative treatment(53.2%), the remainder were treated with surgery alone or surgery with radiation therapy. Information on potential prognostic factors was collected, and then univariate analysis was performed to determine significant prognostic factors. RESULTS: Overall median survival was 7.6 months, with 1-month, 1-year, and 2-year survival rates of 92.4%, 32.9% and 17.9% respectively. Age, performamce status and primary tumor treatment were significant prognostic factors on survival. Number and treatment modality of brain metastases were also identified as prognostic factors. Median survival in patients conservatively treated with steroid alone or with radiotherapy was significantly different from that in patients treated with surgery(p=0031). CONCLUSION: Age(<65), performance status(Karnofsky performance scale>80), number of brain metastases(single), treatmet modalities of primary tumor and brain metastases(surgical treatment) appeared to be statistically significant prognostic factors. These prognostic factors could be helpful to determine subgroups of patients suitable for selection in future trials.