Nutrition is a mature science with well established principles for energy, protein and mineral metabolism based on known metabolic pathways. The quantitative requirements are summarised within various international feeding standards and models. However, when these are applied to specific circumstances, especially in northern Australia, the response of the animal to nutrient supply does not always agree with that predicted from the feeding standards or the error of prediction is not sufficiently accurate for practical use. There is a need for the continual testing of these relationships within production systems. Molecular methods have the potential to discover new metabolic relationships within tissues and characterise the microbial ecology and its relationship to rumen function. Suitable problem models based on growth, meat quality, reproduction, milk and fibre production, and environmental consequences need to be identified. We suggest that production systems designed to meet market weight for age specifications, growth paths and compensatory growth, skeletal growth, parasites, fatty acid isomers, adaptation to low crude protein diets, rumen microbial ecology, epigenetics, remote data acquisition and animal management, greenhouse gas emission, and C balance of various production systems are important problem models, the research of which will benefit the future of the livestock industries in Australia.