This article is designed to explain the importance of medico-economic models for evaluation of the impact of treatments of benign prostatic hyperplasia. The authors illustrate this approach by comparing two hypotheses: transurethral resection of the prostate versus watchful waiting using a Markov type of model. This model is able to simulate the probability of development of events such as transurethral resection of the prostate, acute urinary retention over a 40-year period by 3-month cycles. The model is also able to simulate the mean cost of management of a patient over this period. This type of model may appear slightly artificial and complicated, but it is the only model which could allow long-term simulations of the respective value of the various therapeutic strategies for BPH. This type of model is also evolutive, its performance is gradually improved as new data become available in the literature, allowing refinement of a number of hypotheses.