Many chemopreventives that show efficacy in vitro show little/no response or require bolus doses in animal models and clinical trials because of limited bioavailability. Ellagic acid has been tested in various animal models with mixed results. We report the efficacy of ellagic acid delivered by a subcutaneous implant compared with a dietary route against estrogen-induced mammary tumors. Ellagic acid delayed the first tumor appearance by 2 and 3 weeks by implant and diet routes, respectively. The tumor incidence was 75% and 69% by the implant and dietary routes when the control group had 100% palpable tumors by 26 weeks. Ellagic acid also significantly reduced the tumor burden by both implant (855±242 mm3; P=0.0375) and dietary (599±169 mm3; P=0.0133) routes compared with control (1522±299 mm3). Similar reductions were observed in tumor multiplicity (4.8±0.5; P=0.0042 and 4.5±0.4; P=0.0031 tumors/rat with implant and diet, respectively, vs. 8.9±1.2 in control). The total amount of ellagic acid administered by implant was 5.92±3.48 whereas it was 800±40 mg/rat through diet. Thus, over 130-fold dose reduction produced similar biological responses when delivered by implant. The anticarcinogenicity effects corroborated the observed reduction in levels of pituitary prolactin. This novel approach opens new avenues to test agents individually or as mixtures for their chemopreventive potential that are discontinued, either due to lack of bioavailability or toxicity potentially associated with high doses or due to lack of availability of sufficient quantities.