Nutritional assessment based on animal response factors is the basis of essentially all dietary recommendations. Blood concentrations of nutrients, metabolites, and hormones are important animal response factors associated with nutriture, making blood analysis an important nutritional assessment technique. There are, however, numerous sources of variability, other than nutrition, affecting the concentration of blood analytes used in nutritional assessment. Minimizing the effects of non-nutrient sources of variability and maximizing the effects of nutritional variability is the objective in designing strategies for blood sampling and testing for nutritional assessment. Important non-nutrient sources of variability are age, sex, gestation stage, lactation stage and milk yield, and season. When interpreting test results, grouping animals by these characteristics is an important means of minimizing the effects of non-nutritional variability. Within these groups, it is important to take an adequate number of samples, generally starting out with at least seven. Finally, selecting appropriate tests is critical. Tests commonly used for clinicopathologic evaluations are not necessarily the best tests for nutritional assessment. Analytes should be chosen that are likely to have a large portion of their total variability caused by nutritional effects. This generally does not include those metabolites the blood concentrations of which are rigidly controlled by homeostatic forces.