Several epidemiological studies have associated metformin treatment with a reduction in breast cancer incidence in pre-diabetic and type II diabetic populations. Uncertainty exists regarding which patient populations and/or tumor subtypes will benefit from metformin treatment, and most preclinical in vivo studies have given little attention to the cellular pharmacology of intratumoral metformin uptake. Epidemiological reports consistently link western-style high fat diets, which drive overweight and obesity, with increased risk of breast cancer. We used a rat model of high fat diet (HFD) induced overweight and mammary carcinogenesis to define intratumoral factors that confer metformin sensitivity. Mammary tumors were initiated with N-methyl-N-nitrosourea (MNU), and rats were randomized into metformin-treated (2 mg/ml drinking water) or control groups (water only) for 8 weeks. Two-thirds of existing mammary tumors responded to metformin treatment with decreased tumor volumes (p<0.05), reduced proliferative index (p<0.01), and activated AMPK (p<0.05). Highly responsive tumors accumulated 3-fold greater metformin amounts (p<0.05) that were positively correlated with organic cation transporter-2 (OCT2) protein expression (r=0.57, P=0.038). Importantly, intratumoral metformin concentration negatively associated with tumor volume (P=0.03), and each 10 pmol increase in intratumoral metformin predicted >0.11 cm3 reduction in tumor volume. Metformin treatment also decreased proinflammatory arachidonic acid >1.5 fold in responsive tumors (P=0.023). Collectively, these preclinical data provide evidence for a direct effect of metformin in vivo and suggest that OCT2 expression may predict metformin uptake and tumor response.