Background Information on the cost-effectiveness of malaria control is needed for the WHO Roll Back Malaria campaign, but is sparse. We used mathematical models to calculate cost-effectiveness ratios for the main prevention and treatment interventions in sub-Saharan Africa. Methods We analysed interventions to prevent malaria in childhood (insecticide-treated nets, residual spraying of houses, and chemoprophylaxis) and pregnancy (chloroquine chemoprophylaxis and sulfadoxine-pyrimethamine intermittent treatment), and to improve malaria treatment (improved compliance, improved availability of second-line and third-line drugs, and changes in first-line drug). We developed models that included probabilistic sensitivity analysis to calculate ranges for the cost per disability-adjusted life year (DALY) averted for each intervention in three economic strata. Data were obtained from published and unpublished sources, and consultations with researchers and programme managers. Findings In a very-low-income country, for insecticide treatment of existing nets, the cost-effectiveness range was US$4-10 per DALY averted; for provision of nets and insecticide treatment $19-85; for residual spraying (two rounds per year) $32-58; for chemoprophylaxis for children $3-12 (assuming an existing delivery system); for intermittent treatment of pregnant women $4-29; and for improvement in case management $1-8. Although some interventions are inexpensive, achieving high coverage with an intervention to prevent childhood malaria would use a high proportion of current health-care expenditure. Interpretation Cost-effective interventions are available. A package of interventions:to decrease the bulk of the malaria burden is not, however, affordable in very-low-income countries. Coverage of the most vulnerable groups in Africa will require substantial assistance from external donors.