Findings from basic nutrition research poorly transfer to humans, and thus, the results are “lost in translation.” The objective was to identify a mathematical model that accurately translates diet and nutritional bioactives between humans and preclinical rodent models. A secondary objective was to provide guidance to researchers and reviewers for appropriate and inappropriate experimental designs when translational fidelity for dietary bioactives/nutrients between species is desired or inferred. This manuscript reviews the standard preclinical experimental rodent diets originally designed by the America Institute of Nutrition (AIN), and reviews the most common allometric scaling models, their strengths and weaknesses, for the extrapolation of nutrients between species. This includes extrapolations based on body weight, surface area and metabolic rate, and caloric needs of each of the species. To investigate the translational fidelity of each mathematical model, all of the micronutrients and the essential fatty acids in each of the AIN diets were converted to a human equivalent dose, and these values were compared to the recommended or actual intakes in the US population. The use a mathematical scaling model based on differences in caloric needs between species more accurately reflects recommendations in humans and would enhance translational validity and minimize false-positive results.