OBJECTIVE: To document the cost incurred by the Tanzanian government by changing the policy on first-line treatment of malaria, from chloroquine to sulfadoxine-pyrimethamine. METHODS: Costs were analysed from the perspective of the Ministry of Health and included all sources of funding. Costs external to the public health sector (e.g. private and community costs) were not included. The base case analysis adopted an incremental rather than a full cost approach, assuming that an organizational infrastructure was already in place. However, specific attention was paid to the burden placed on National Malaria Control Program staff. We also costed activities planned but not implemented to estimate the total expense for an 'ideal' process. RESULTS: Total costs were Tsh 795 million (USD 813,743), with the largest proportion accounted for by training. Costs of the policy change process were equivalent to about 4% of annual government and donor expenditure on malaria and to about 1% of overall public expenditure on health. A number of planned activities were not implemented; including these would bring the total cost to Ts 880 million (USD 896,130). CONCLUSION: On top of extra costs for the drugs themselves, a change in treatment policy requires time, resources and substantial management capacity at national and local level. A better understanding of these issues and the costs involved benefits countries planning and implementing policy change.