BACKGROUND:The role of clinicopathologic characteristics of the recurrent tumor in determining survival in a cohort of patients with ipsilateral breast tumor recurrence (IBTR) was investigated. METHODS: Among 6020 women with pT1-T2, pN0-1, M0 treated with breast-conserving surgery from 1989 to 1999, 269 developed isolated IBTR. Ten-year Kaplan-Meier breast cancer-specific survival (BCSS) and overall survival (OS), calculated from date of IBTR, were analyzed according to clinicopathologic characteristics. RESULTS: Factors that were associated with diminished OS and BCSS on univariate analysis were: time to IBTR 48 months, lymphovascular invasion positive status, estrogen receptor (ER) negative status, high grade, clinical IBTR detection, biopsy alone, and close/positive margins (all P < .05). On multivariate analysis, time to IBTR 48 months (hazard ratio [HR], 1.87, P ¼ .012), lymphovascular invasion positive status (HR, 2.46; P < .001), ER negative status (HR, 2.08; P ¼ .013), high-grade recurrent disease (HR, 1.88; P ¼ .013), and close/positive margins after surgery for IBTR (HR, 1.94; P ¼ .013) retained significance for prediction of diminished OS. When stratified according to number of adverse prognostic features, 10-year OS was 70.4% in patients with 1 factor, 35.8% with 2 factors, and 19.9% with 3 or more factors (P < .001). CONCLUSIONS: Time to recurrence 48 months, lymphovascular invasion positive status, ER negative status, high-grade histology, and close/positive margins in association with the recurrent tumor are independent prognostic factors for survival after IBTR. The presence of 2 or more of these features at recurrence is significantly associated with poor OS. These criteria can be used to prognosticate and guide clinical decisions after recurrence.