Ninety-five women who underwent blind contralateral breast biopsy during surgical treatment of a known breast cancer primary were studied prospectively. All biopsies were performed between 1981 and 1989. Patients with palpable or mammographic abnormalities prompting the contralateral biopsy were excluded so that the study sample included only truly blind contralateral biopsies. Only two infiltrating carcinomas were found, resulting in a positive biopsy rate of 2.1% for invasive disease. Three additional biopsies showed only lobular carcinoma in situ, a finding that usually does not alter clinical management. One patient with a negative contralateral biopsy developed invasive carcinoma in that breast within 2 years of the biopsy. The authors were unable to identify any subgroup of patients at increased risk of a positive contralateral biopsy. These results suggest that blind biopsy of the contralateral breast performed at the time of the initial treatment of breast carcinoma is not an efficient method of cancer detection. Alternative management strategies are discussed.