We present an economic decision support model, based on a Bayesian network, for Mycoplasma infection in slaughter swine production. The model describes the various risk factors for Mycoplasma infection and their interactions. This leads to a stochastic determination of the consequences of productivity factors, and this again results in stochastic economic consequences when changing the risk factors or engaging in control arrangements. We use the model to calculate how estimated prevalence and level of Mycoplasma changes when we gather evidence from various veterinarian examinations, and we show how this influences the distribution of the economic results of a change in strategy to fight Mycoplasma. One result is that the profitability in fighting Mycoplasma alone is questionable, i.e. risky, thus we must consider the loss of all related diseases in a decision context.