Results from numerous trials have indicated that breast‐conserving therapy (BCT) produces outcomes equivalent to those produced by mastectomy in terms of both locoregional control and survival. However, conservative treatment has resulted in the dilemma of how best to address recurrences when they appear in a breast treated previously with radiation therapy. Attempts have been made to characterize ipsilateral breast tumor recurrences (IBTRs) as either true recurrences of the treated malignancy or new primary carcinomas, because cancers that represent new primary tumors may be associated with a more favorable prognosis compared with cancers that represent true recurrences.The authors studied the clonality of IBTRs relative to the initial invasive carcinomas by using a polymerase chain reaction loss‐of‐heterozygosity molecular comparison assay in 29 patients who received breast‐conserving therapy (BCT).Twenty‐two IBTRs (76%) were related clonally to the initial carcinoma, and 7 IBTRs (24%) were clonally different. Clonally related IBTRs were more frequently higher grade (72.2% vs 14.3%; P = .009) and developed sooner after initial treatment (mean time to IBTR, 4.04 years in clonally related IBTRs vs 9.25 years in clonally different IBTRs; P = .002). Six patients subsequently developed distant metastases, and 5 of those patients (83.3%) had clonally related IBTRs. Clinical IBTR classification and molecular clonality assay results differed in 30% of all patients. The proportion of IBTRs that were related clonally at 5 years, 10 years, and 15 years after BCT were 93%, 67%, and 33%, respectively.Clinical classifications of IBTRs were unreliable methods for determining clonality in many patients. Molecular clonality assays provided a reliable means of identifying patients who may benefit from aggressive systemic therapy at the time of IBTR and also provided a more accurate assessment of the efficacy of various forms of local therapy. Cancer 2007. © 2007 American Cancer Society. [ABSTRACT FROM AUTHOR]