The complexity and importance of agricultural operations management has increased as agriculture have adapted capital-intensive production systems stimulating the development of more formal planning techniques. Planning tools being adapted include decision support systems (DSS) aimed at achieving better production control. Specifically, this project addresses operational planning or scheduling of field operations. This planning procedure involves the allocation of time and resources to activities considering inherent uncertainty and risks. The method used is applied decision analysis together with the principles of Bayesian networks. Bayesian networks capture the relations (which may be uncertain or imprecise) between a set of random variables associated with some domain. By augmenting the network with decision variables and a utility function, an influence diagram is created. This combined approach enables the formal description of a decision problem involving conditional, functional and informational dependencies. The DSS is built by using an expert system shell HUGIN (Handling Uncertainty by General Influence Networks) providing the knowledge representation framework and inference engine necessary. Crop condition, weather forecasts and labour/machinery states are specified together with their mutual dependencies and conditional probabilities. Given some prior information and specific observations the model will update all probabilities. For example, in the, case of managing harvesting operations, the model will give the probable number of harvesting hours, probable state of the crop with regard to quantity and quality, when given evidence on current moisture content, weather prognosis, availability of labour and machinery etc.