OBJECTIVE: The primary objective of the analysis was to simultaneously account for and characterize the uncertainty in a cost-effectiveness model of current triptan (5-HT agonist) therapy for migraine headache. METHODS: Common components of cost and effect were identified as migraine incidence, overall market share for each triptan, migraine response rate at two hours per triptan [milligram] strength, migraine recurrence rate per triptan strength, and the percentage of patients on each triptan strength. Components specific to cost alone were the number of migraine attacks/patient/month and wholesale acquisition cost (WAC) per triptan strength. Effect was defined as a successfully treated migraine attack. Costs were determined for every triptan and not per triptan strength. Uncertainty in the cost-effectiveness ratio was modeled using Monte Carlo simulation techniques, specifically utilizing inverse transformation of the component cumulative distribution functions. One thousand random draws were taken from each of the assumed component distributions, and cost and effect were computed for each draw. Cost, effect, and their random components were log-10 transformed, and ordinary least squares regression was performed using cost/effect as the dependent variable and the components of cost and effect as independent variables. Distributional assumptions were logistic-normal for proportional components, Poisson for the attacks/patient/month component, and uniform for the WAC component. RESULTS: There was a 10% change in the cost-effectiveness ratio with a 10% change in attacks/patient/month and WAC, while the change in the ratio was less than 8% with a 10% change in the response and recurrence rates. All other components were insignificant in their effect. CONCLUSIONS: The cost-effectiveness model appears to be stable under the assumptions of its construction, since cost/effect did not change at a rate greater than the rate of change for any of the significant components of cost and effect.