Curcumin is widely known for its antioxidant, anti-inflammatory, and antiproliferative activities in cell-culture studies. However, poor oral bioavailability limited its efficacy in animal and clinical studies. Recently, we developed polymeric curcumin implants that circumvent oral bioavailability issues, and tested their potential against 17β-estradiol (E2)-mediated mammary tumorigenesis. Female Augustus Copenhagen Irish (ACI) rats were administered curcumin either via diet (1,000 ppm) or via polymeric curcumin implants (two 2 cm; 200 mg each; 20% drug load) 4 days before grafting a subcutaneous E2 silastic implant (1.2 cm, 9 mg E2). Curcumin implants were changed after 4.5 months to provide higher curcumin dose at the appearance of palpable tumors. The animals were euthanized after 3 weeks, 3 months, and after the tumor incidence reached >80% (~6 months) in control animals. The curcumin administered via implants resulted in significant reduction in both the tumor multiplicity (2 ± 1 vs. 5 ± 3; P = 0.001) and tumor volume (184 ± 198 mm(3) vs. 280 ± 141 mm(3); P = 0.0283); the dietary curcumin, however, was ineffective. Dietary curcumin increased hepatic CYP1A and CYP1B1 activities without any effect on CYP3A4 activity, whereas curcumin implants increased both CYP1A and CYP3A4 activities but decreased CYP1B1 activity in the presence of E2. Because CYP1A and CYP3A4 metabolize most of the E2 to its noncarcinogenic 2-OH metabolite, and CYP1B1 produces potentially carcinogenic 4-OH metabolite, favorable modulation of these CYPs via systemically delivered curcumin could be one of the potential mechanisms. The analysis of plasma and liver by high-performance liquid chromatography showed substantially higher curcumin levels via implants versus the dietary route despite substantially higher dose administered.