Three thousand three hundred three women were followed an average of 17 years following benign breast biopsy. These women comprise 84.4% of those originally targeted for follow-up. Risk of invasive breast cancer development was analyzed in relation to the hyperplasia classification scheme of Wellings et al (JNCI 1975; 55:231-73) that is based on terminal ductal-lobular units (ALA). Cancer risk was also assessed with respect to family history of breast cancer in first-degree relatives (FHBC), as well as atypical features of hyperplasia recognized by resemblance to carcinoma in situ of ductal type (ADH). There was a trend of increasing cancer risk with increasing degree of ALA lesion, reaching 1.9 with ALA-IV lesions having both qualitative and quantitative features of advanced atypical hyperplasia. When ADH lesions are removed from the analysis, any predictive power of ALA lesions is lost. ADH recognizes histologic lesions with a four- to fivefold increased risk of breast cancer. FHBC interacts with any hyperplastic lesion so as to approximately double the cancer risk.