While hormone-dependent, mammary tumors induced with carcinogens (DMBA or NMU) in intact rats have been used extensively for studying aromatase inhibitors, there is currently no suitable model to investigate their effects in human breast cancers in vivo. While hormone responsive tumors can be formed in the athymic mouse using human breast carcinoma MCF-7 cells, due to the low ovarian estrogen production, tumor growth is induced with estradiol supplementation. Thus, this model is unsuitable for studies of aromatase inhibitors. We have induced tumors without the need for estrogen supplementation by co-inoculating MCF-7 cells with Matrigel, a basement membrane preparation, into intact athymic mice. In one experiment, 45 days after inoculation, mice were assigned to the control group or 4-hydroxyandrostenedione (4-OHA) (1 mg/day s.c.) treatment for 52 days. Tumor volumes in the control mice increased 672%, whereas tumor volumes in the treated mice did not change significantly (178.9 +/- 16.2 to 336.6 +/- 120 mm3). In the second experiment, 55 days after inoculation, groups of mice were treated with the antiestrogen, tamoxifen (5 micrograms/day s.c.) or vehicle (controls). Tumor volumes in the control mice increased 325% in 58 days, whereas there was no significant change in tumor volume in the tamoxifen treated group (338.8 +/- 55.3 to 330.6 +/- 84.9 mm3). The results suggest that (1) the tumors resulting from MCF-7 cells co-inoculated with Matrigel are estrogen-dependent and (2) tamoxifen and 4-OHA were effective in suppressing growth of these tumors. The results suggest that this model should be useful for evaluating the effects of aromatase inhibitors and for comparing breast cancer treatments.