531 Background: There remains variation in the use of radiation (RT) in women with DCIS, despite prospective randomized trials documenting its benefit in reducing the risk of ipsilateral breast tumor recurrence (IBTR). Methods: Patients with DCIS treated with excision alone or excision plus radiation at our institution from 1991–1995 were identified. A dedicated breast pathologist assessed margin width, number of involved ducts at closest margin, nuclear grade, and presence of necrosis and lobular neoplasia. Pathological variables, age, and presence of palpable mass were tested in uni- and multivariate analysis for association with risk of IBTR and added value of RT. Results: 294 patients with a median follow-up of 11 years had actuarial 10- and 15-year overall IBTR rates of 22 and 29%. For lesions excised with margins of <1mm, 1–9mm, and ≥10mm, the actuarial 10-year IBTR rates were 28, 21, and 19%, respectively. RT reduced adjusted IBTR rates by 62% (12 vs. 28%; p=0.002) for all patients; 83% for lesions with <1mm margins (p=0.002), 70% for 1–9mm (p=0.05), and 27% (p=0.55) for ≥10mm. Following adjustment for other variables, higher volume of disease near the margin was associated with risk of IBTR in the no RT group (HR=3.37, p=0.002) and greater benefit of RT (HR 0.14; p=0.004). Conclusions: Effect of RT on local recurrence risk is influenced by both margin width and number of involved ducts at nearest margin. Patients with higher volume of disease near the margin derive a greater benefit from the addition of RT. Despite margins of ≥10mm, the risk of local recurrence remains substantial in patients with DCIS. [Table: see text] No significant financial relationships to disclose.