In order to resolve the question of which ovarian steroid stimulates normal human mammary epithelial cell proliferation, we have implanted pieces of normal human breast tissue subcutaneously into athymic nude mice. These mice were then treated with slow-release pellets containing estradiol (E2) or progesterone (P) such that serum levels of E2 and P were increased to those seen in normal women. The proliferative activity of the tissue implants was assessed by uptake of tritiated thymidine and steroid receptor expression was measured immunocytochemically. Insertion of a 2 mg E2 pellet 14 days after tissue implantation increased the thymidine labeling index (TLI) from a median of 0.4% (n = 34) to a median of 2.1% after 7 days (n = 43; P < 0.001 by Mann Whitney U test). In contrast, treatment with a P pellet (4 mg) had no effect upon the TLI whereas P (4 mg) in combination with E2 (2 mg) had no effect over and above that of E2 alone. There was a significant correlation between the increase in TLI and either the E2 content of the pellets (P < 0.001 by linear regression) or the serum E2 levels achieved (P < 0.001). Expression of the P receptor was increased 15- to 20-fold by E2 treatment. We conclude that E2 is sufficient to stimulate human breast epithelial cell proliferation at physiologically relevant concentrations and that P does not affect proliferation either alone or after E2 priming.