Exact knowledge of the nutrient requirements of premature infants is critically important for the prevention of postnatal growth failure and for improved neurodevelopmental outcome. Methods whereby nutrient requirements can be estimated fall into two categories, factorial methods and empirical methods. Each have their advantages and disadvantages. The factorial methods provide estimates of requirements for protein, energy and a number of other nutrients. The exact methods used can vary but still yield fairly similar results. Factorial methods also permit estimation of the extra nutrients needed for a given degree of catch-up growth, but cannot indicate the extent to which catch-up growth is actually possible. Empirical methods yield estimates of the requirements for protein and energy but not for other nutrients. They often give an indication of what degree of catch-up growth is possible, in addition to providing estimates of the requirements for protein and energy. The advantages of catch-up growth outweigh any possible disadvantages associated with it. The nutrients needed for catch-up growth should therefore always be provided.