OBJECTIVES: To calculate the NNT to avoid a recurrence of GIST after resection, to compare the cost of adjuvant treatment with imatinib (IM) with the cost of recurrence, and to estimate a budget impact from the Brazilian Public Health Care System (SUS) perspective. METHODS: Available relative risk reduction at 1 year from the Z9001 clinical trial  and historical rate of recurrence for no adjuvant treatment were applied to estimate absolute risk reduction and NNT. Adjuvant treatment effect was extrapolated to 3-year period as ongoing trials are investigating longer treatment duration (SSG XVIII). A 5-year time horizon was set for cost comparison and Budget Impact Analysis (BIA). Incremental Cost to Avoid Recurrence (ICAR) was defined as the difference between the cost of adjuvant treatment (IM, monitoring) and the cost of recurrence (IM, surgery, monitoring, best supportive care). ICAR was applied to adjuvant GIST incidence for BIA. Epidemiological data  (incidence, proportion of resectables); health access, diagnosis and expected adjuvant treatment rates were obtained from literature. Resource utilization  and cost data came from official guideline and administrative databases, literature and expert opinion. Costs are reported in 2007 Euros. A 5% discount rate was applied. Univariate sensitivity analysis was performed. RESULTS: The NNT to avoid a recurrence was estimated at 2.1 based on extrapolated GIST recurrence risk profile in Brazil. Cost of adjuvant treatment was €50,298 and the cost of a recurrence €61,998. Annual ICAR was €8,725. The annual impact on the Ministry of Health budget was 0.01%, which included impact on infrastructure (e.g. monitoring costs) of SUS. Results were sensitive to the recurrence rate and adjuvant treatment duration. CONCLUSIONS: Considering that imatinib is already reimbursed by SUS for metastatic/unresectable GIST, adjuvant therapy for primary GIST represents good value for money for the prevention of recurrence, and an annual budget impact of 0.01%.