This paper describes a model for forecasting the treatment costs for hydrocephalic patients with ventriculoperitoneal shunts. Modeling with institution-specific or reported failure rates allows the prediction of shunt failure in real and/or theoretical populations. The addition of costing factors (derived from hospitalization, operative and drug costs) to the model allows the derivation of partial or total cost estimates. The effects of varying the failure rate, infection rate, number of new patients, number of lost patients and costing factors can be simulated and measured. Basing this model on data from our institution, decreasing the rate of failure during the first year following shunt insertion or revision has the potential for greater cost savings over time than either decreasing the shunt infection rates or the duration of hospital stay. By combining shunt performance and financial data, an estimate of the cost of the treatment of a population with hydrocephalus, over time, can be derived. These data can be critical for institutional and program budgeting and serve as an estimate of the economic effects of treatment changes proposed in clinical trials.