OBJECTIVES: To assess the cost-effectiveness of nasal spray zolmitriptan 5 mg (ZOL), compared to subcutaneous sumatriptan 6 mg (SUM) in the management of moderate to severe migraine patients in Belgium. METHODS: A medical decision analytic model was developed in MS-Excel, reflecting patient outcomes and related management. The net response rate (headache response after 2 hours) above placebo is a primary outcome in migraine trials and was obtained from randomised clinical trials. This response rate above placebo was 52% for SUM, and 39.7% for ZOL. However, recurrence rates after initial response and within the same attack were 39% and only 25.6% respectively. Cost of non-response and cost of managing recurrence were obtained from published local literature from a health care payer perspective. Full response was defined as patients responding without recurrence. The time horizon was limited to one attack episode. RESULTS: ZOL total treatment cost was €28.02 with 43.7% full response, while SUM costed €40.95 in total with 42.19% full response (i.e. ZOL slightly dominant). Rank order stability analyses (ROSA) showed that results were very robust towards variations in cost of management (max. deviation in savings of 4%). The analysis was sensitive to treatment performance and recurrence rates, both having a weak effect on savings (max deviation in savings of 13%), but with the potential to inverse the dominant position of ZOL. CONCLUSIONS: ZOL as standard treatment in moderate to severe migraine is cost saving from the Belgian health care payer's perspective compared to SUM, with comparable effectiveness.