Background: Phase II trials serve to provide safety and efficacy data for evaluating novel interventions that may be selected for large Phase III trials. To help define benchmarks for efficacy, we evaluated outcomes of all recently published metastatic breast cancer Phase II trials (MBP2). Methods: All English language MBP2 published from June 2005 through June 2010 were identified by Pubmed search. Trial characteristics (type, line, regimen), tumor type, and outcomes were analyzed. Line of therapy was categorized as first-line (first line only), early-line (2^nd and 1-2 line), and any/late-line (2+ lines or not specified). Descriptive statistics, Wilcoxon signed rank, and univariate linear regression were performed. Results: 226 Phase II trials were abstracted. 31% were first-line, 30% early-line, and 39% any/late-line. 83% (187) were single-arm design; 12% (27) and 5% (11) were randomized and cohort studies, respectively. Overall response rate (ORR) was reported in 204 (90%) MBP2, stable disease (SD) in 127 (56%), time to progression (TTP) in 185 (82%), progression free survival in 20 (9%), and overall survival (OS) in 143 (63%). SD was defined as 6 months in 29%, but definitions varied, and SD was not defined in 43%. Overall, median ORR was 36% (IQR: 16-52, range 0-98%); TTP was 27 weeks (IQR: 15-36, range 4-122), and OS was 65 weeks (IQR: 52-95). For first-line, median ORR, TTP, and OS were 53%, 35 weeks, and 97 weeks, respectively. In first-line, ORR ranged from 5%-98%, TTP from 9-72 weeks, and OS from 13-884 weeks. For early-line vs. any/late line, median ORR was 30% vs. 22%, median TTP was 24 vs. 19 weeks, and median OS was 60 vs. 56 weeks. ORR, TTP, and OS differed significantly by line of therapy (all p<0.0001). Univariate linear regression demonstrated a significant association between ORR, TTP, and OS (p<0.005). Conclusions: Appraisal of recent trials can provide benchmarks for evaluation of novel therapy. As with standard interventions, outcomes in MBP2 vary by line of therapy. Standardization in reporting outcomes, particularly definition of SD, would improve assessment and support rational decision making for which interventions should move forward to Phase III trials.