Application of nutritional knowledge generally requires expression in mathematical form. It is insufficient to conclude that animals should be fed more of a nutrient; the recommendation must include an estimate of how much more. Thus, models that accurately and precisely represent animal responses to varying nutrient supply are a critical product of nutrition science. They also allow quantitative hypothesis testing which guides the scientific process. Hence, the construction of models and the derivation of parameter estimates for those models are a critical component of nutrition science. For this learning exercise, it is assumed the participant is proficient in the use of R and the use of linear and nonlinear regression functions, and has participated in the National Animal Nutrition Program Level 1 Workshop, which includes a module on building a portion of this model, or has gained that expertise through self-study. Participants will be given a data set and a model containing 4 components that interact. These are (1) a pool of insoluble N in the rumen, (2) a pool of soluble N in the rumen, (3) a pool of ruminal microbes that are consuming soluble N, and (4) digestion and absorption of N from the small intestine. The model will be fitted to observed ruminal and fecal N outflow data to derive parameter estimates for the conversion of insoluble N to soluble N and the fractional use of the soluble N in support of microbial growth, and digestion of N in the intestines. The exercise demonstrates the use of an optimizer to fit model parameters to observed data.