OBJECTIVES: Acute Myeloid Leukaemia (AML) can be diagnosed at any age and accounts for approximately one third of all leukaemia diagnoses. Treatment, which can be given with supportive and/or curative intent, is considered expensive compared to other cancers. Despite this, no long-term predictive models have been developed for AML, mainly due to the complexities associated with this disease. Thus, the aim of the current study was to develop an AML model (based on a UK cohort) that would allow cost and life expectancy results to be expressed at population level. METHODS: The model developed in this study combined a decision tree with several Markov models. This was in order to reflect the complexity of health states, treatments, and prognostic factors (such as age and response to chemotherapy) of AML. The model was simulated over a life cycle of 60 months and results were contrasted between two age-groups and over different treatment pathways. Probabilistic modelling was also implemented in order to capture the potential uncertainties of input parameters. Transition probabilities, life expectancies and costs were derived from the NHS Hospital Episode Statistics (HES) and a UK population-based database from the Haematological Malignancy Research Network (HMRN, www.hmrn.org). RESULTS: The expected five year medical cost and life expectancy for the elderly patients (>60) were £22,538 and 8.5 months respectively, and for the young adult patients (18-60) £82,266 and 33.78 months respectively. The model was validated by the fact that the predicted results captured 92% of the actual costs, while it also demonstrated good fit of the actual survival outcomes. CONCLUSIONS: Costs and life expectancy of AML vary according to patient characteristics and treatment pathways. It is expected that future application of the AML model developed in this study could be used to evaluate new diagnostic tools/treatments and to support healthcare decision makers.