Abstract: Objective To establish gene assays for determining (anti) estrogen effect of environmental chemicals; and to compare the reactivity and sensitivity of two assays with different estrogen subtype. Methods Human estrogen receptor α (hERα) and hERβ mediated reporter gene assays employing firefly luciferase (Luc) were developed. The expression plasmid hERα or hERβ was constructed and transiently co-transfected into LLC-MK2 cells with pERE-minP-Luc2P reporter plasmid and the control plasmid pGL4.74. Estradiol (E 2) and diethylstilbestrol (DES) served as positive test substances to verify the performance of the assays. The effectiveness of the assays for detecting anti-estrogenic activity was tested using 10-5 mol/L ICI 182,780 under different concentrations of E 2. The performance of the two subtype-mediated assays was verified and compared using bisphenol A (BPA) and genistein (GS). Results The hERα mediated assay found expression of reported gene at 1.9×10-11mol/L E 2; and the largest luciferase activity was shown at 10-8 mol/L E 2, resulting in 30.7-fold of vehicle control. The hERβ mediated assay found expression of reporter gene at 2.2×10-11 mol/L E 2, and the largest luciferase activity was shown at 10-8 mol/L E 2, resulting in 14.4-fold of vehicle control. ICI 182,780 inhibited estrogenic activity of E 2 significantly. In both assays, E 2 failed to induce luciferase activity without hER-pcDNA3.1. BPA and GS induced luciferase activity. Conclusion Both assays have high sensitivity and reproducibility for detecting (anti) estrogen effect. The pGL4-based hERβ has lower sensitivity than the hERα-mediated reporter gene assay. BPA shows stronger estrogenic activity than GS in hERα mediated reporter gene assay; whereas, GS shows stronger estrogenic activity than BPA in hERβ mediated reporter gene assay.