In the last 6 years a number of non-randomized, predominantly single institutional trials of breast conservation therapy (BCT) with DCIS, have demonstrated that it constitutes a very heterogeneous group of diseases with markedly different risks of local recurrence and invasive transformation. There has been a consensus that DCIS, which exhibits a "comedo" morphology, generally defines a high risk group. Most studies, moreover, have identified the same two features, nuclear grade and necrosis, as contributing most significantly to prognosis. Nuclear grade and necrosis have been identified as independent prognostic variables in several studies. High nuclear grade DCIS which exhibits comedo necrosis defines the majority of all DCIS which will result in local recurrence and invasive transformation after BCT. Studies utilizing image cytometry, to determine ploidy and S-phase fraction and immunohistochemical studies of proliferation and oncogene distribution have shown a significant association with morphologically identified high nuclear grade and aneuploidy, high S-phase fraction or proliferation rate, presence of HER-2/neu and P53 oncogenes and absence of estrogen receptors. Generally the inverse of this association is seen with low nuclear grade DCIS. However, initial hopes that these adjunctive studies would identify subsets within the high nuclear grade group which might be more likely to recur have not been fulfilled.