The high concentrations of estradiol (E2) found in breast tumors of postmenopausal women could be the result of enhanced uptake from plasma or in situ aromatization of androgens to estrogens. To test the relative importance of these two mechanisms, a model system allowing precise distinction between each is required. Such a model was established using aromatase (A+)- and sham (A-)-transfected MCF-7 cells inoculated into ovariectomized (OVX) nude mice. To validate the model, the confounding effect of peripheral aromatization was first excluded experimentally. A- cells were inoculated into OVX mice as homoimplants (A- cells on both flanks) or heteroimplants (A- cells on one flank and A+ cells on the other), and growth of A- cells in response to exogenous aromatase substrate, androstenedione (delta4A), was evaluated. A- cells did not grow in either group during the 8 weeks of observation, indicating the lack of peripheral aromatization in OVX mice. The biological effects of in situ aromatization were then directly examined. We found that A+ cells in the heteroimplant group grew rapidly, and that the average weight of A+ tumor was 7.6-fold larger and tissue E2 concentration was 3-4-fold higher than A- tumors grown in the same animals. These results demonstrate that in situ aromatization rather than uptake can be a determinant of tumor E2 content and growth stimulation. An additional experiment was then designed to evaluate the relative importance of in situ synthesis versus uptake under conditions reflecting postmenopausal physiology. Groups of OVX mice bearing A+ cells received E2 Silastic implants to clamp plasma levels at 5, 7, 10, and 20 pg/ml or delta4A by injection. The highest tumor E2 concentration and growth rate were found in the group receiving delta4A. E2 delivered by Silastic implants always produced lower tissue E2 levels and tumor growth rates than resulted from in situ synthesis. These data provide direct evidence that under physiological conditions reflecting those in postmenopausal women, in situ aromatization in breast tumor makes a major contribution to tissue E2 content. As further validation that our experimental paradigm models the postmenopausal state, we studied OVX animals not given delta4A as substrate. A+ cells also grew under these conditions, and the aromatase inhibitor 4-hydroxyandrostenedione reduced both tumor E2 level and growth rate, providing additional evidence of the importance of in situ synthesis. These studies provide the first direct evidence that in situ synthesis of E2 in breast tumors, as opposed to peripheral aromatization and uptake from plasma, can enhance tissue E2 levels and stimulate tumor growth.