OBJECTIVES:  To compare expected costs per AML patient in remission using decision tree and cost-in-use analysis based on a comparative 10-country phase III clinical trial. METHODS: Using a five-year time horizon, costs were estimated from the payor’s perspective for patients in their first complete remission from AML.  Clinical endpoints were remission (Leukemia Free), relapse and death.  Resources consumed were taken from clinical trial data and supplemented with cost information from the literature and advisors (for patients in relapse).  Comparators were histamine dihydrochloride + low dose interleukin-2 (n= 129) vs. standard of care (n=132).  Unit costs were taken from UK sources including NHS reference cost, British National Formulary 56, National Blood Services and the literature for concomitant medications, blood products, emergency room visits, physician visits and relapses.  Cost for interleukin was included; however, the investigated drug cost was not included in the analysis as no price has been set to date.  We computed the expected cost by treatment for each method, using a 5% discount rate. RESULTS: Overall 5 year Leukemia Free Survival for treatment vs. standard of care was 2.23 years vs. 1.75 (P=0.02), respectively.  Expected costs per treatment arm for the tree method, treatment vs standard of care, were £40,725 vs. £39,371, respectively, while for the cost-in-use method, treatment vs. standard of care was, £40,209 vs. £41,702, respectively.  The tree method overestimated expected cost for the treatment arm by 1.3%, and underestimated the cost for standard of care by 5.6%. CONCLUSIONS: The two methods estimated similar values. However, the cost-in-use method yields a more accurate estimate compared to the tree method because the tree method does not adjust for events that take place between nodes, thus possibly introducing error.  The cost-in-use method captures resources with known time points, minimizing over- or under-reporting of resources consumed.