Using the example of bovine viral diarrhoea (BVD) in Scottish suckler (cow-calf) beef herds, this paper demonstrated a method to establish the maximum average net benefit of disease control under specific epidemiological and farm business circumstances. Data were generated for the method using a stochastic epidemiological model set to estimate the mean and variance of control costs and output losses from BVD for 50-cow or 120-cow herds, either free of BVD at the outset or of unknown BVD status. Control of disease was by increased investment in a variety of ('biosecurity') measures aimed at reducing the probability of virus entering the closed herd in any 1 year of a 10-year period of simulated exposure to risk from BVD virus introduction either with or without vaccination. Herds free of BVD at the outset enjoyed much greater maximum average net benefits than herds of unknown BVD status. Best allocations of hypothetical incentives to encourage farmers to establish their freedom from BVD were therefore outlined. Vaccination and biosecurity were generally found to be complementary rather than substitutes for one another. The advantages of the maximum net benefit measure over the more usual average total cost of endemic disease were demonstrated and discussed. The maximum net benefit method focuses on the relationship between costs and benefits, which often exhibits diminishing marginal returns meaning that profit maximisation and disease minimisation are incompatible. The method can also allow for constraints on and competition for limited farm resources. It was argued that these attributes are important to persuade farmers to invest in animal health.