The 2-phenylindole system has been identified as a suitable structure for the design of non-steroidal pure estrogen antagonists [E. von Angerer et al., J. Steroid Biochem. Molec. Biol. 49 (1994) 51-62]. Derivatives with an amide function in the side chain antagonized the stimulatory effect of estrogens both in vitro and in vivo, and showed no agonistic activity when given alone. The findings of other groups who studied steroidal antiestrogens prompted us to replace the amide function by sulfide, sulfoxide, sulfone, sulfonamide and related groups. The compounds with polar sulfur functions retained the high binding affinity for the calf uterine estrogen receptor (RBA: 1-5% of estradiol; ICI 182,780; 6.2%). The estrogenic effect was quantified in a transcription assay using HeLa cells cotransfected with the expression vector HEG0 for the human estrogen receptor and a reporter plasmid that harbored a Vit. A2 ERE and the luciferase gene driven by a thymidine kinase promotor. Pentylsulfide, -sulfinyl, and -sulfonyl groups, linked to the indole nitrogen by a decamethylene spacer, were devoid of any transcriptional activity. These results were confirmed in the mouse uterine weight test. The sulfone (ZK 164,015) completely abolished the effect of a standard dose of estrone at a daily dose of 7 mg/kg. This compound strongly inhibited the growth of hormone-sensitive human MCF-7 breast cancer cells with an IC50-value close to 1 nM. Similar activity was found for the steroidal sulfoxide ICI 182,780. We were also able to demonstrate significant antineoplastic activity in vivo for some of these new 2-phenylindole derivatives.