Making critical decisions in a relatively short period of time involving various biophysical and economic factors having potential impact on business is an integral part of activities of farmers including those wool-producing sheep growers in the Western Australian Sheep Zone. A biophysical simulation model representing annual pasture system, grazed by sheep, was used to estimate the effect of various combination of grazing management practices on biological outputs (i.e., wool production, sheep liveweight, and pasture production) in a wide range of climatic, biological, and management conditions. The dynamic programming models were used to estimate the optimal grazing management practices in various circumstances (i.e., wool prices, feed supplement prices, discount rates, and risk attitudes). Optimal stocking rate is related to the initial pasture density. Pasture deferment plays an important role in the management of annual pasture system. Results from the risk-neutral stochastic model are not radically different to those from the deterministic model. The key differences are that optimal stocking rate and maximum net returns are slightly lower than in the deterministic model.