PURPOSE Estrogen-related receptor alpha (ERRα) signaling has recently been implicated in breast cancer. We investigated the clinical value of ERRα in randomized cohorts of tamoxifen-treated and adjuvant-untreated patients. EXPERIMENTAL DESIGN Cox proportional hazards regression was used to evaluate the significance of associations between ERRα gene expression levels and patient DMFS in a previously published microarray dataset representing 2,000 breast tumor cases derived from multiple medical centers worldwide. The 912 tumors used for immunostaining were from a tamoxifen-randomized primary breast cancer trial conducted in Stockholm, Sweden, during 1976-1990. Mouse model was used to study the effect of tamoxifen treatment on lung colonization of MDA-MB-231 control cells and MDA-MB-231 cells with stable knockdown of ERRα. The phenotypic effects associated with ERRα modulation were studied using immunoblotting analyses and wound-healing assay. RESULTS We found that in ER-negative and triple-negative breast cancer (TNBC) adjuvant-untreated patients, ERRα expression indicated worse prognosis and correlated with poor outcome predictors. However, in tamoxifen-treated patients, an improved outcome was observed with high ERRα gene and protein expression. Reduced ERRα expression was oncogenic in the presence of tamoxifen, measured by in vitro proliferation and migration assays and in vivo metastasis studies. CONCLUSIONS Taken together, these data show that ERRα expression predicts response to tamoxifen treatment, and ERRα could be a biomarker of tamoxifen sensitivity and a prognostic factor in TNBC.