Originally designed as an antitumor agent, zidovudine (AZT) has exhibited only marginal tumor growth inhibitory activity. Recently, three abstracts have described positive clinical outcomes for a small number of patients with advanced breast cancer treated with weekly infusions of either methotrexate or cisplatin and AZT. Consequently, we conducted a preclinical study of the anti-breast cancer and anti-mammary tumor activity of AZT. Here we have demonstrated that AZT, alone, has a preferential in vitro and in vivo effect on breast and mammary cancer cells. It is 1000 times as potent as an inhibitor of the in vitro growth of the human breast cancer cell line MCF-7 (IC50 = 10 +/- 5 nM) than of the growth of the T-cell leukemia cell line CEM (IC50 = 14 +/- 2 microM). A novel mechanism for this preferential effect on growth is indicated by the 3-4-fold increase in production of phosphorylated AZT (mono-, di-, and triphosphate) in MCF-7 relative to CEM. We extended these in vitro observations to in vivo studies in rats and found that AZT is a potent in vivo inhibitor of the growth of methylnitrosourea-induced rat mammary tumors without any apparent toxic effects on internal organs. These preclinical results demonstrate, for the first time, that AZT has significant anti-breast cancer activity and strongly suggest that the clinical usefulness of this drug is worthy of investigation.