IN HER 3 OCTOBER News Features “The littlest patient”(p. 24) and “Hope in a mouse”(p. 28), J. Couzin-Frankel discusses the utility of mouse genetic cancer models for testing the efficacy of chemotherapeutic drugs [also discussed in (1)]. She omits the importance of mouse diet. Standard chow diets vary between batches in both macro and micro nutrients (2). These diets can elevate biochemically powerful phytoestrogens to mouse serum levels 50,000 times as high as those of endogenous estrogen (3). Standard chow diets also provide high levels of vitamin D3, a hormone with wide-ranging effects on genes and pathways affecting growth and differentiation in multiple tissues (4). As a result, vitamin D serum levels are elevated (5, 6) to levels far exceeding the range in the US human population (7). Yet, vitamin D signaling in the stroma profoundly influences chemotherapeutic efficacy for pancreatic cancer (8). Many other nutrients also substantially alter tumor development and phenotype. It seems foolish to believe that even the most elegant mouse genetic models will be highly useful without paying attention to nutrient intake. Newmark and Lipkin conducted pioneering work on rodent diets that reflect nutrient intake common in populations at risk for tumor development (9–11). Unfortunately, these lessons and a tremendous body of literature in epidemiology, carcinogenesis, and chemoprevention are routinely ignored in high-profile work on the molecular biology and genetics of cancer.