Among 41,109 women diagnosed with breast cancer between 1935 and 1982 in Connecticut, 3,984 developed a second cancer, whereas 2,426 were expected [relative risk (RR) = 1.64; 95% CI = 1.6-1.7]. This increased risk persisted for 30 years and was highest in women under 55 years of age at the time of breast cancer diagnosis. Second primary breast cancers (RR = 3.0) accounted for almost one-half of all new neoplasms. However, if subsequent breast cancers were excluded, the risk for all other second cancers was only 1.15 (95% CI = 1.10-1.20), and no excess risk was seen among women over age 55 at initial breast cancer. Significant risks were found for cancers of the ovary (RR = 1.7) and uterine corpus (RR = 1.4), possibly linked with shared reproductive factors such as nulliparity or late age at menopause. Malignant melanoma (RR = 1.5), thyroid cancer (RR = 1.6), and colon cancer (RR = 1.2) were also significantly elevated; possible shared risk factors remain to be elucidated. Significant deficits of multiple myeloma and chronic lymphocytic leukemia were noted. Women who received initial radiotherapy compared with those who did not were at slightly higher risk of developing a second cancer, most notably acute nonlymphocytic leukemia, non-Hodgkin's lymphoma, and cancers of the esophagus, kidney, and connective tissue, although the nature of the associations was not always clear. Some of the soft tissue sarcomas were lymphangiosarcomas of the arm, a consequence of the lymphedema that may complicate radical mastectomy (Stewart-Treves syndrome). Women treated with radiation were at higher risk of developing a second breast neoplasm (RR = 3.9) than nonirradiated women (RR = 2.8). Further investigation should focus on the mechanisms underlying the relationships between breast, genital tract, and colon cancers, and on the effects of treatment modalities on the risk of subsequent neoplasms.