634 Background: Despite the widespread perception that the availability of new drugs improves prognosis in MBC, survival gains were rarely reported in randomised clinical trials of 1st line chemotherapy (CTX) with last generation agents. We evaluated the survival of MBC patients enrolled into trials conducted by the GONO during a 20 year period, to detect any improvement potentially attributable to the use of newer drugs. Methods: Between 1983 and 2001, 790 patients were entered into 6 consecutive trials of CTX coordinated by the data center at the National Cancer Research Institute in Genoa, Italy (2 phase II and 4 randomised phase III). In trials started after 1994, a taxane was part of the CTX regimen. This analysis was limited to 640 patients, as patients metastatic ab initio were excluded. Five cohorts of patients according to 3-yr periods of diagnosis of relapse were identified. Kaplan-Meier estimates of overall survival (OS) and time to progression (PFS) were computed from enrolment, and compared by the log rank test. Multivariate PH models were used to adjust for differences in baseline factors, including age, nodal and receptor status, adjuvant CTX, relapse free survival, metastatic site and PS. Results: The following differences in baseline characteristics in the 5 cohorts were observed: the proportion of patients pretreatead with adjuvant anthracyclines increased from 4% to 26%; that of patients with a RFS>5 years increased from 19% to 33% and the proportion of patients with dominant visceral disease increased from 57% to 74%. Median PFS was 306 days, 294 days, 237 days, 328 days and 348 days in cohorts 1983–86, 87–89, 92–94, 95–97 and 98–2001, respectively (p < 0.0001). Median OS in the same cohorts was 546 days, 522 days, 584 days, 793 days and 713 days, respectively (p < 0.0001). These data were confirmed in multivariate analyses, when prognostic variables were adjusted for. Conclusions: It appears that the prognosis of MBC patients enrolled in CTX clinical trials since 1995 is significantly improved as compared to patients enrolled during 1983–94. A possible role of drugs introduced in the '90s is suggested. No significant financial relationships to disclose.