The purpose of this study is to determine the prognostic role of Ki67 evaluated in relapse biopsies from patients with metastatic breast cancer (MBC). Two hundred and ten patients diagnosed with MBC in Stockholm, Sweden between 1998 and 2009 and with Ki67 assessed at time of first systemic relapse (mKi67) were retrospectively identified and divided into two groups according to mKi67 fraction (low ≤20 %, high >20 %). Post-relapse survival was compared between the groups using Kaplan-Meier and Cox regression methods. Death rate as function of continuous mKi67 was also evaluated. Furthermore, the prognostic role of intra-individual change in Ki67 between primary tumor and matched metastasis was explored by Kaplan-Meier plots. One hundred and twenty-five patients had low and 85 had high mKi67. Median survival was 25 and 17 months in low- and high-mKi67 group, respectively [hazard ratio (HR) 0.69, 95 % confidence intervals (CI) 0.51-0.92, P = 0.01]. In a multivariate model adjusted for prognostic confounders, low-mKi67 showed a non-significant trend toward better survival (HR 0.85, 95 %CI 0.62-1.16, P = 0.30). Nevertheless, mKi67 independently correlated with survival when compared with primary tumor proliferation (HR 0.56, 95 %CI 0.38-0.81, P = 0.002). The 2-year death rate steeply increased as mKi67 increased. Moreover, the change from high in primary tumor to low in metastasis significantly correlated with longer survival when compared with stable Ki67 levels (HR 0.48, 95 %CI 0.31-0.76, P = 0.002). In this cohort of MBC patients, mKi67 inversely but not independently correlated with survival. However, a significant association between mKi67 and survival was shown regardless of primary tumor proliferation.