Recent policy recommendations have called for increased research efforts to inform the design of cost-effective interventions to address the shortage of health workers in rural areas. This paper takes forward the recent use of Discrete Choice Experiments to assess the effects of potential incentives to attract nurses to rural areas. The analysis relies on data collected in South Africa between August and November 2008. Effectiveness measures derived from Discrete Choice Experiments are combined in a Markov model to derive the long-term effects of policies, and costs are evaluated with secondary data. Measures involving the selection of more nursing students who are more likely to accept positions in rural areas are shown to be the most cost-effective interventions. If such policies could not be implemented, the next best options are to offer preferential access to specialist training to nurses willing to work in rural areas.