Animal simulation models have been developed to aid in understanding how different factors are related to beef animal performance. The utility of these models becomes evident when one considers the many relationships and interrelationships that occur in grazing systems. The choice of an appropriate animal model must be made before grazing systems can be simulated. This paper evaluates the adequacy of a stocker steer model in predicting live weight gain, average daily gain and daily dry matter intake. A rumen kinetics model and a temperature stress component were incorporated to take differences in forage quality and environmental temperature into account. Digestion and other parameters were determined from literature values or by calibrating the model with actual intake data. The model was used to simulate three years of an actual grazing study. Addition of the rumen model resulted in significant improvement of the steer model. While steer live weight was sensitive to the added temperature stress logic, addition of the temperature stress component did not significantly improve the steer model. In two of the three years, the model slightly overestimated live weight. Predicted live weights at the end of the grazing seasons ranged from 6·4 kg less, to 14·1 kg more, than observed weights.