A combined database of 342 patients with DCIS treated by lumpectomy alone versus lumpectomy and radiation therapy with a median 82-month follow-up is summarized in this joint study. Reproducible subtype classification and common methods of mammographic-pathologic correlation and complete tissue processing are unique features of this database, and they permit outcome to be analyzed by pathologic subtype, size, and margine status. Striking differences are noted in local control rates analyzed by subtype, which were largely independent of irradiation (see Table 1). Analysis of local recurrence-free survival restricted to those cases with a 10 mm or larger free margin width revealed no significant differences between the irradiated and nonirradiated groups. The local recurrence rates were 5% in those treated by lumpectomy alone and 4.5% in those treated by lumpectomy and irradiation (Table 4). Although differences in local recurrence rates for DCIS with a 10 mm plus free margin, with or without irradiation, were noted, they were not large. For DCIS patients with adequate (10 mm or more) or intermediate (1-9 mm) margin width, there was a reduction in local recurrence limited to the high-grade subtype (group III) with radiation therapy; an absolute 8% reduction for those with adequate margins and 11% for those with intermediate margins, but the difference was significant only for the latter group (Table 5). However, no significant differences were noted for the lower grade DCIS subtypes (groups I and II). For DCIS patients with inadequate margins (i.e., less than 1 mm), irradiation provided no benefit for local control. We conclude that an adequate surgical excision for DCIS, defined as a free margin of 10 mm or more, largely makes moot the question of local control related to pathologic subtype and treatment modality. Specifically, adequately excised high-grade (group III) DCIS received a benefit for local control from radiation therapy of only 8% within the median follow-up period. This difference is not significant. The impact of DCIS size or extent on local recurrence is much smaller than margin width (see Table 3). Significant differences achieved by radiation therapy were demonstrable only for the smallest size group (15 mm or less) in the highgrade subtype (group III). Differences in local recurrence rates for low and intermediate subtypes (group I and II) based on radiation therapy could not be demonstrated within the three size categories used in the study. We conclude that although adequate margins are more difficult to achieve for larger or more extensive DCIS, size alone is not a prohibition to breast conservation.