89 Background: Ki-67 expression in breast cancer has been described as predictive of pathological complete response and prognostic of recurrence free survival (RFS). Our aim was to evaluate in a retrospective cohort of triple negative breast cancer patients if tumor proliferation measured by ki-67 expression is correlated with the outcome. Methods: We evaluated a retrospective cohort of 109 cases of triple negative breast cancer (ER-, PR- and HER2- determined by immunohistochemistry). Ki-67 labeling index was determined in Formalin-Fixed, Paraffin-Embedded residual tumors after neoadjuvant chemotherapy. Patients were stratified in Ki67<15% and Ki67≥15%. Clinicopathological data was retrieved from clinical records. RFS and overall survival (OS) were calculated using the Kaplan–Meier method and variables compared using the log-rank or Breslow test and Hazard Ratios (HR) estimated by the Cox regression. Results: The median age for patients was 47.5 years, 55 (50.5%) were premenopausal and 54 (49.5%) postmenopausal. Eight patients (7.3%) were clinical stages II and 101 (92.7%) stages III. Median of Ki-67 expression was 35.97% (0.96% - 77.7%). There was not association between Ki-67 expression (<15% VS ≥15%) with tumor size, nodal involvement, clinical stage and menopausal status. After a median of follow of 21.6 months, 62 patients (56.9%) have relapsed and 53 (48.6%) have die. The median time for RFS and OS were 21.2 and 31.4 months, respectively. Median of RFS was 12.6 months for Ki67<15% vs 21.2 months for Ki67≥15%, P=0.421 (HR=0.91). Median of OS was 34.9 months for Ki67<15% vs 31.4 months for Ki67≥15%, P=0.755 (HR=1.18). Only nodal involvement was found predictor of shorter RFS (0 nodes, 25.2 months vs 1 -3 nodes, 26.1 months, vs >3 nodes, 9.4 months, P=0.020). Conclusions: Ki-67 labeling index was not related with the outcome in terms of OS and RFS in patients with residual triple negative breast tumors that were treated with neoadjuvant chemotherapy.