Concern over the appearance of algal blooms in Summer months on Australian waterways has been increasing in recent years. Some algae are toxic to humans and livestock when ingested. To avoid the dangers of contaminated drinking water, costs must be incurred, either by eliminating nutrient loadings, or taking action to disperse blooms. Such costs may also be worth incurring to prevent the loss of recreation use of waterways and consequent loss of tourism revenue. The aim is to consider the formulation of models to aid in the control of outbreaks of algal blooms, so as to maximise the net present value of social net returns. The complexity of determining the triggering of blooms is discussed. Results from regressing algal cell counts on possible explantory variables such as water flow, water temperature, and nutrient loadings for three sites on the River Murray are presented. Dynamic programming models are formulated with phosphorus in sediments, algal cell count, and water in storage as alternative state variables. The role of rainfall is an important stochastic variable is considered, because flood events lead to substantial nutrient runnoff, and dry periods are associated with low river flow and conditions favourable for the outbreak of blooms.