The effects of a single 17 beta-estradiol (E2) injection on cell proliferation were studied in 3 groups of 30 mice transplanted with the MXT ovary-dependent mammary tumor. In group A, all animals were castrated prior to tumor implantation; groups B and C had intact ovaries at the time of transplantation, but group B was left intact throughout the experiment, while group C underwent castration 4 weeks later. On day 40 in groups B and C and on day 80 in group A, in which tumor development was significantly delayed, the same procedure for testing the effects of E2 was applied: Ten controls received 0.1 ml saline ip and were killed on the next day; 4 lots of 5 mice received 0.25 micrograms E2 ip and were killed one by one at 12-hour intervals. Exactly 1 hour prior to sacrifice, each animal received 25 microCi [methyl-3H]thymidine ip. Histologic sections of tumors and uteri were processed for autoradiography, and nuclear thymidine (dThd) labeling indices (LI) were determined. All tumors of group A grafted under unfavorable hormonal conditions were poorly differentiated, and E2 injection induced no appreciable changes in their dThd LI. Tumors B and C were well-differentiated adenocarcinomas, in which E2 induced significant modifications of cell proliferation. In group B, complex changes in dThd LI occurred in tumors as well as in uteri, probably due to interferences with the ovarian hormonal production. In group C, E2 produced a marked rise in dThd LI in tumors, lasting from the 12th to the 36th hour after its injection. Stimulation was maximum at the 24th hour, representing a 2.8-fold increase over mean basal dThd LI. It is concluded that the presence of an intact ovarian function at the time of transplantation is critical for maintaining the properties of hormone dependence in MXT tumors. In mice castrated after tumor implantation, a single E2 injection induces a marked and partially synchronous proliferation of neoplastic cells. The hypothesis that such hormonal manipulation might amplify the killing effect of cell cycle- or phase-specific cytotoxic drugs could be adequately tested with this model.