Histoculture drug response assay (HDRA) systems have been used in evaluating the cytotoxic effects of chemotherapeutic agents for many kinds of advanced cancers. We have adapted the HDRA system to estimate the antitumor effect of aromatase (estrogen synthetase) inhibitors on breast cancer. Small pieces of breast cancer tissue specimens were placed onto a collagen-matrix filled with medium containing testosterone (a substrate for aromatase) or testosterone plus an aromatase inhibitor. At the end of culture, [3H]-thymidine incorporation was measured in aliquots of the histocultured specimens after 10 days culture. The increment of thymidine incorporation in testosterone-treated specimens to that of control provides an index of existence of aromatase and estrogen-dependency, since converted estradiol from added testosterone by aromatase stimulates the incorporation. The decrease in the index of "testosterone + aromatase inhibitor"/"testosterone" indicates the antitumor effect of the aromatase inhibitor on breast cancer. Twenty-one 25 breast cancer surgical specimens were successfully cultured, and 6 showed the increased incorporation of [3H]-thymidine by testosterone. Aromatase inhibitor blocked this stimulation in these 6 specimens. These results suggested that this antitumor effect is related to the inhibition of aromatase and the aromatase inhibitor would be effective for individual patients with breast cancer which responds to testosterone in this histoculture assay system. The histoculture technique we used here is therefore expected to be useful in predicting the efficacy of aromatase inhibitors for individual patients with breast cancer.